diff --git a/adapter_model.bin b/adapter_model.bin
index 1b6fe766ab6e3bde1e95c1168e8b27f515e22af0..b595165cbaf9902530b4e987ef92ffae7b616d47 100644
--- a/adapter_model.bin
+++ b/adapter_model.bin
@@ -1,3 +1,3 @@
version https://git-lfs.github.com/spec/v1
-oid sha256:e7b701342ea311c1e25f0f6953505f048df3f56027dc5c764b95df4491bfde1c
+oid sha256:22fbdcaf7857de1b74e03d99a156e9d6d448734b321467b134a7651acfe0c5fa
size 39409357
diff --git a/checkpoint-100/README.md b/checkpoint-100/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3547d02a238ad8dbfc3eb1872f34cbdb49334c82
--- /dev/null
+++ b/checkpoint-100/README.md
@@ -0,0 +1,138 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+### Framework versions
+
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+
+- PEFT 0.6.0.dev0
diff --git a/checkpoint-100/adapter_config.json b/checkpoint-100/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c661f74e40394c079de68f07861d73c341803b67
--- /dev/null
+++ b/checkpoint-100/adapter_config.json
@@ -0,0 +1,25 @@
+{
+ "alpha_pattern": {},
+ "auto_mapping": null,
+ "base_model_name_or_path": "bigscience/bloomz-3b",
+ "bias": "none",
+ "fan_in_fan_out": false,
+ "inference_mode": true,
+ "init_lora_weights": true,
+ "layers_pattern": null,
+ "layers_to_transform": null,
+ "lora_alpha": 16,
+ "lora_dropout": 0.0,
+ "modules_to_save": null,
+ "peft_type": "LORA",
+ "r": 8,
+ "rank_pattern": {},
+ "revision": null,
+ "target_modules": [
+ "dense_4h_to_h",
+ "dense",
+ "dense_h_to_4h",
+ "query_key_value"
+ ],
+ "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/checkpoint-100/adapter_model.bin b/checkpoint-100/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..5935b1162e1aa3c4c7d7f3bd55d1953230d1e7c6
--- /dev/null
+++ b/checkpoint-100/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8335990cc53e924dd84264d3c9f79ff9c558d5b3398b3e77273713338fe9152d
+size 39409357
diff --git a/checkpoint-100/optimizer.pt b/checkpoint-100/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..057600cb206f898b07e7f00702de1976c5c8b750
--- /dev/null
+++ b/checkpoint-100/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bace7c30bbad8abd3be9019a0d69fc2cab14aeca0e48fcb0e5ae051d6fdc579
+size 78844421
diff --git a/checkpoint-100/rng_state.pth b/checkpoint-100/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e8a62beafdd6e61c41abbc54c8b7428f4e3484c0
--- /dev/null
+++ b/checkpoint-100/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2278a87cdf86c3f9219223c847f6b27f6b7f15b8226b617f38936e8ff2cbcde
+size 14575
diff --git a/checkpoint-100/scheduler.pt b/checkpoint-100/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d91eabecece4023a37656bfcb691d472b7f9c489
--- /dev/null
+++ b/checkpoint-100/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1640c3b85caa3e129cded850309149bec1a2e020090ed853140950215c11127
+size 627
diff --git a/checkpoint-100/special_tokens_map.json b/checkpoint-100/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdafe480f024ff444c7492147536765ce5d55a2d
--- /dev/null
+++ b/checkpoint-100/special_tokens_map.json
@@ -0,0 +1,6 @@
+{
+ "bos_token": "",
+ "eos_token": "",
+ "pad_token": "",
+ "unk_token": ""
+}
diff --git a/checkpoint-100/tokenizer.json b/checkpoint-100/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..673c31abdeadf6576c3c754df86459e1ad64e207
--- /dev/null
+++ b/checkpoint-100/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85b00d7db4df5df2e3f01cacc3feda246002a672f3356eec7f4b04a22eb0dfbe
+size 14500570
diff --git a/checkpoint-100/tokenizer_config.json b/checkpoint-100/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b56cc9c2965c07132c35df3e2972e93d98c82c3
--- /dev/null
+++ b/checkpoint-100/tokenizer_config.json
@@ -0,0 +1,10 @@
+{
+ "add_prefix_space": false,
+ "bos_token": "",
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "",
+ "model_max_length": 1000000000000000019884624838656,
+ "pad_token": "",
+ "tokenizer_class": "BloomTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoint-100/trainer_state.json b/checkpoint-100/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..21653c01843acad0db276e3b9f185feee6d33c5a
--- /dev/null
+++ b/checkpoint-100/trainer_state.json
@@ -0,0 +1,619 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 0.5586592178770949,
+ "eval_steps": 500,
+ "global_step": 100,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019972067039106145,
+ "loss": 2.6443,
+ "step": 1
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019944134078212292,
+ "loss": 2.4104,
+ "step": 2
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019916201117318435,
+ "loss": 2.4975,
+ "step": 3
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019888268156424582,
+ "loss": 2.3513,
+ "step": 4
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0001986033519553073,
+ "loss": 2.4274,
+ "step": 5
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.00019832402234636873,
+ "loss": 2.3628,
+ "step": 6
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0001980446927374302,
+ "loss": 2.3567,
+ "step": 7
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.00019776536312849163,
+ "loss": 2.4121,
+ "step": 8
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.00019748603351955307,
+ "loss": 2.4033,
+ "step": 9
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019720670391061454,
+ "loss": 2.2805,
+ "step": 10
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019692737430167598,
+ "loss": 2.2639,
+ "step": 11
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019664804469273744,
+ "loss": 2.2724,
+ "step": 12
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019636871508379888,
+ "loss": 2.332,
+ "step": 13
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019608938547486035,
+ "loss": 2.2261,
+ "step": 14
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019581005586592182,
+ "loss": 2.2208,
+ "step": 15
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.00019553072625698326,
+ "loss": 2.3351,
+ "step": 16
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0001952513966480447,
+ "loss": 2.2475,
+ "step": 17
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.00019497206703910616,
+ "loss": 2.3283,
+ "step": 18
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0001946927374301676,
+ "loss": 2.1346,
+ "step": 19
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.00019441340782122907,
+ "loss": 2.131,
+ "step": 20
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0001941340782122905,
+ "loss": 2.1718,
+ "step": 21
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.00019385474860335195,
+ "loss": 2.2446,
+ "step": 22
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0001935754189944134,
+ "loss": 2.306,
+ "step": 23
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.00019329608938547488,
+ "loss": 2.1908,
+ "step": 24
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.00019301675977653632,
+ "loss": 2.2844,
+ "step": 25
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019273743016759779,
+ "loss": 2.2235,
+ "step": 26
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019245810055865922,
+ "loss": 2.1842,
+ "step": 27
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019217877094972066,
+ "loss": 2.2675,
+ "step": 28
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019189944134078213,
+ "loss": 2.2532,
+ "step": 29
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019162011173184357,
+ "loss": 2.1788,
+ "step": 30
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019134078212290504,
+ "loss": 2.2494,
+ "step": 31
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0001910614525139665,
+ "loss": 2.1995,
+ "step": 32
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.00019078212290502794,
+ "loss": 2.1451,
+ "step": 33
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0001905027932960894,
+ "loss": 2.223,
+ "step": 34
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.00019022346368715085,
+ "loss": 2.2854,
+ "step": 35
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0001899441340782123,
+ "loss": 2.2265,
+ "step": 36
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.00018966480446927375,
+ "loss": 2.1214,
+ "step": 37
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0001893854748603352,
+ "loss": 2.1898,
+ "step": 38
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.00018910614525139666,
+ "loss": 2.1974,
+ "step": 39
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0001888268156424581,
+ "loss": 2.2259,
+ "step": 40
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018854748603351957,
+ "loss": 2.2094,
+ "step": 41
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018826815642458103,
+ "loss": 2.1731,
+ "step": 42
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.00018798882681564247,
+ "loss": 2.2373,
+ "step": 43
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0001877094972067039,
+ "loss": 2.2295,
+ "step": 44
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.00018743016759776538,
+ "loss": 2.1947,
+ "step": 45
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018715083798882682,
+ "loss": 2.2115,
+ "step": 46
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018687150837988828,
+ "loss": 2.1224,
+ "step": 47
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018659217877094972,
+ "loss": 2.2137,
+ "step": 48
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018631284916201116,
+ "loss": 2.2338,
+ "step": 49
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.00018603351955307266,
+ "loss": 2.1298,
+ "step": 50
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0001857541899441341,
+ "loss": 2.0883,
+ "step": 51
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.00018547486033519553,
+ "loss": 2.1216,
+ "step": 52
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.000185195530726257,
+ "loss": 2.2112,
+ "step": 53
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.00018491620111731844,
+ "loss": 2.1224,
+ "step": 54
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0001846368715083799,
+ "loss": 2.2375,
+ "step": 55
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.00018435754189944135,
+ "loss": 2.2235,
+ "step": 56
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018407821229050279,
+ "loss": 2.1682,
+ "step": 57
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018379888268156425,
+ "loss": 2.2077,
+ "step": 58
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.00018351955307262572,
+ "loss": 2.1596,
+ "step": 59
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018324022346368716,
+ "loss": 2.1311,
+ "step": 60
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018296089385474862,
+ "loss": 2.1333,
+ "step": 61
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018268156424581006,
+ "loss": 2.0901,
+ "step": 62
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018240223463687153,
+ "loss": 2.1971,
+ "step": 63
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.00018212290502793297,
+ "loss": 2.2602,
+ "step": 64
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0001818435754189944,
+ "loss": 2.2194,
+ "step": 65
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018156424581005588,
+ "loss": 2.1218,
+ "step": 66
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018128491620111731,
+ "loss": 2.2049,
+ "step": 67
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.00018100558659217878,
+ "loss": 2.1521,
+ "step": 68
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.00018072625698324025,
+ "loss": 2.112,
+ "step": 69
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0001804469273743017,
+ "loss": 2.1906,
+ "step": 70
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.00018016759776536313,
+ "loss": 2.1717,
+ "step": 71
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0001798882681564246,
+ "loss": 2.0712,
+ "step": 72
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.00017960893854748603,
+ "loss": 2.141,
+ "step": 73
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0001793296089385475,
+ "loss": 2.0656,
+ "step": 74
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017905027932960894,
+ "loss": 2.1125,
+ "step": 75
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017877094972067038,
+ "loss": 2.0869,
+ "step": 76
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.00017849162011173187,
+ "loss": 2.2478,
+ "step": 77
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0001782122905027933,
+ "loss": 2.1535,
+ "step": 78
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.00017793296089385475,
+ "loss": 2.1927,
+ "step": 79
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017765363128491622,
+ "loss": 2.1213,
+ "step": 80
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017737430167597766,
+ "loss": 2.0981,
+ "step": 81
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017709497206703912,
+ "loss": 2.1828,
+ "step": 82
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017681564245810056,
+ "loss": 2.0562,
+ "step": 83
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.000176536312849162,
+ "loss": 2.1334,
+ "step": 84
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.00017625698324022347,
+ "loss": 2.1225,
+ "step": 85
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.00017597765363128493,
+ "loss": 2.2098,
+ "step": 86
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017569832402234637,
+ "loss": 2.1519,
+ "step": 87
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017541899441340784,
+ "loss": 2.1132,
+ "step": 88
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017513966480446928,
+ "loss": 2.0333,
+ "step": 89
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017486033519553075,
+ "loss": 2.2764,
+ "step": 90
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017458100558659218,
+ "loss": 2.1838,
+ "step": 91
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017430167597765362,
+ "loss": 2.1386,
+ "step": 92
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0001740223463687151,
+ "loss": 2.1034,
+ "step": 93
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.00017374301675977656,
+ "loss": 2.0346,
+ "step": 94
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.000173463687150838,
+ "loss": 2.0274,
+ "step": 95
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.00017318435754189946,
+ "loss": 2.1036,
+ "step": 96
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0001729050279329609,
+ "loss": 2.1208,
+ "step": 97
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.00017262569832402237,
+ "loss": 2.0572,
+ "step": 98
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0001723463687150838,
+ "loss": 2.1702,
+ "step": 99
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.00017206703910614525,
+ "loss": 2.1302,
+ "step": 100
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 716,
+ "num_train_epochs": 4,
+ "save_steps": 100,
+ "total_flos": 5.117136720224256e+16,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-100/training_args.bin b/checkpoint-100/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8672c716e925d0028b4938db147703f58656ff7
--- /dev/null
+++ b/checkpoint-100/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:188ae1c421cc0c6435d1f71d8d3423ac4abc7dba0e6fc2efcbc4dbe77c741317
+size 4027
diff --git a/checkpoint-200/README.md b/checkpoint-200/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..2f257a448caef1c59022426ad3ec9ba80acb3821
--- /dev/null
+++ b/checkpoint-200/README.md
@@ -0,0 +1,60 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+### Framework versions
+
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+
+- PEFT 0.6.0.dev0
diff --git a/checkpoint-200/adapter_config.json b/checkpoint-200/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c661f74e40394c079de68f07861d73c341803b67
--- /dev/null
+++ b/checkpoint-200/adapter_config.json
@@ -0,0 +1,25 @@
+{
+ "alpha_pattern": {},
+ "auto_mapping": null,
+ "base_model_name_or_path": "bigscience/bloomz-3b",
+ "bias": "none",
+ "fan_in_fan_out": false,
+ "inference_mode": true,
+ "init_lora_weights": true,
+ "layers_pattern": null,
+ "layers_to_transform": null,
+ "lora_alpha": 16,
+ "lora_dropout": 0.0,
+ "modules_to_save": null,
+ "peft_type": "LORA",
+ "r": 8,
+ "rank_pattern": {},
+ "revision": null,
+ "target_modules": [
+ "dense_4h_to_h",
+ "dense",
+ "dense_h_to_4h",
+ "query_key_value"
+ ],
+ "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/checkpoint-200/adapter_model.bin b/checkpoint-200/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..099d9fc77ee96c743e6855b11d9098013e8ca3a9
--- /dev/null
+++ b/checkpoint-200/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5d07783590970cc01440af02efb30baf84be013f42d6e324dbec3f1211d5743
+size 39409357
diff --git a/checkpoint-200/optimizer.pt b/checkpoint-200/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..15140fee1abc06c78da052b519728e1113c22927
--- /dev/null
+++ b/checkpoint-200/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bb4e74f21744cbd194fc93b7a41891675165b22a0068ad1db19aeac49015587
+size 78844421
diff --git a/checkpoint-200/rng_state.pth b/checkpoint-200/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b9111227e3b79b9bc3e2a642832c3e49e36216d4
--- /dev/null
+++ b/checkpoint-200/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:972139d83957a9cf2600cb6eeca17287d7a5377c33a53500ae7e13fe830ad36b
+size 14575
diff --git a/checkpoint-200/scheduler.pt b/checkpoint-200/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b1a8820c8482d068bfc50bf5d18b879546282775
--- /dev/null
+++ b/checkpoint-200/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49d3e390b4500e3448050a2123ac7e837c199dcd6015f166fda34e0372e32849
+size 627
diff --git a/checkpoint-200/special_tokens_map.json b/checkpoint-200/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdafe480f024ff444c7492147536765ce5d55a2d
--- /dev/null
+++ b/checkpoint-200/special_tokens_map.json
@@ -0,0 +1,6 @@
+{
+ "bos_token": "",
+ "eos_token": "",
+ "pad_token": "",
+ "unk_token": ""
+}
diff --git a/checkpoint-200/tokenizer.json b/checkpoint-200/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..673c31abdeadf6576c3c754df86459e1ad64e207
--- /dev/null
+++ b/checkpoint-200/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85b00d7db4df5df2e3f01cacc3feda246002a672f3356eec7f4b04a22eb0dfbe
+size 14500570
diff --git a/checkpoint-200/tokenizer_config.json b/checkpoint-200/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b56cc9c2965c07132c35df3e2972e93d98c82c3
--- /dev/null
+++ b/checkpoint-200/tokenizer_config.json
@@ -0,0 +1,10 @@
+{
+ "add_prefix_space": false,
+ "bos_token": "",
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "",
+ "model_max_length": 1000000000000000019884624838656,
+ "pad_token": "",
+ "tokenizer_class": "BloomTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoint-200/trainer_state.json b/checkpoint-200/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..180bc505d2b0627db89719f80c131cfebb1dccad
--- /dev/null
+++ b/checkpoint-200/trainer_state.json
@@ -0,0 +1,1219 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 1.1173184357541899,
+ "eval_steps": 500,
+ "global_step": 200,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019972067039106145,
+ "loss": 2.6443,
+ "step": 1
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019944134078212292,
+ "loss": 2.4104,
+ "step": 2
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019916201117318435,
+ "loss": 2.4975,
+ "step": 3
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019888268156424582,
+ "loss": 2.3513,
+ "step": 4
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0001986033519553073,
+ "loss": 2.4274,
+ "step": 5
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.00019832402234636873,
+ "loss": 2.3628,
+ "step": 6
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0001980446927374302,
+ "loss": 2.3567,
+ "step": 7
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.00019776536312849163,
+ "loss": 2.4121,
+ "step": 8
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.00019748603351955307,
+ "loss": 2.4033,
+ "step": 9
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019720670391061454,
+ "loss": 2.2805,
+ "step": 10
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019692737430167598,
+ "loss": 2.2639,
+ "step": 11
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019664804469273744,
+ "loss": 2.2724,
+ "step": 12
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019636871508379888,
+ "loss": 2.332,
+ "step": 13
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019608938547486035,
+ "loss": 2.2261,
+ "step": 14
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019581005586592182,
+ "loss": 2.2208,
+ "step": 15
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.00019553072625698326,
+ "loss": 2.3351,
+ "step": 16
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0001952513966480447,
+ "loss": 2.2475,
+ "step": 17
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.00019497206703910616,
+ "loss": 2.3283,
+ "step": 18
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0001946927374301676,
+ "loss": 2.1346,
+ "step": 19
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.00019441340782122907,
+ "loss": 2.131,
+ "step": 20
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0001941340782122905,
+ "loss": 2.1718,
+ "step": 21
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.00019385474860335195,
+ "loss": 2.2446,
+ "step": 22
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0001935754189944134,
+ "loss": 2.306,
+ "step": 23
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.00019329608938547488,
+ "loss": 2.1908,
+ "step": 24
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.00019301675977653632,
+ "loss": 2.2844,
+ "step": 25
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019273743016759779,
+ "loss": 2.2235,
+ "step": 26
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019245810055865922,
+ "loss": 2.1842,
+ "step": 27
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019217877094972066,
+ "loss": 2.2675,
+ "step": 28
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019189944134078213,
+ "loss": 2.2532,
+ "step": 29
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019162011173184357,
+ "loss": 2.1788,
+ "step": 30
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019134078212290504,
+ "loss": 2.2494,
+ "step": 31
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0001910614525139665,
+ "loss": 2.1995,
+ "step": 32
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.00019078212290502794,
+ "loss": 2.1451,
+ "step": 33
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0001905027932960894,
+ "loss": 2.223,
+ "step": 34
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.00019022346368715085,
+ "loss": 2.2854,
+ "step": 35
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0001899441340782123,
+ "loss": 2.2265,
+ "step": 36
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.00018966480446927375,
+ "loss": 2.1214,
+ "step": 37
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0001893854748603352,
+ "loss": 2.1898,
+ "step": 38
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.00018910614525139666,
+ "loss": 2.1974,
+ "step": 39
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0001888268156424581,
+ "loss": 2.2259,
+ "step": 40
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018854748603351957,
+ "loss": 2.2094,
+ "step": 41
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018826815642458103,
+ "loss": 2.1731,
+ "step": 42
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.00018798882681564247,
+ "loss": 2.2373,
+ "step": 43
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0001877094972067039,
+ "loss": 2.2295,
+ "step": 44
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.00018743016759776538,
+ "loss": 2.1947,
+ "step": 45
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018715083798882682,
+ "loss": 2.2115,
+ "step": 46
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018687150837988828,
+ "loss": 2.1224,
+ "step": 47
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018659217877094972,
+ "loss": 2.2137,
+ "step": 48
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018631284916201116,
+ "loss": 2.2338,
+ "step": 49
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.00018603351955307266,
+ "loss": 2.1298,
+ "step": 50
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0001857541899441341,
+ "loss": 2.0883,
+ "step": 51
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.00018547486033519553,
+ "loss": 2.1216,
+ "step": 52
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.000185195530726257,
+ "loss": 2.2112,
+ "step": 53
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.00018491620111731844,
+ "loss": 2.1224,
+ "step": 54
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0001846368715083799,
+ "loss": 2.2375,
+ "step": 55
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.00018435754189944135,
+ "loss": 2.2235,
+ "step": 56
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018407821229050279,
+ "loss": 2.1682,
+ "step": 57
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018379888268156425,
+ "loss": 2.2077,
+ "step": 58
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.00018351955307262572,
+ "loss": 2.1596,
+ "step": 59
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018324022346368716,
+ "loss": 2.1311,
+ "step": 60
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018296089385474862,
+ "loss": 2.1333,
+ "step": 61
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018268156424581006,
+ "loss": 2.0901,
+ "step": 62
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018240223463687153,
+ "loss": 2.1971,
+ "step": 63
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.00018212290502793297,
+ "loss": 2.2602,
+ "step": 64
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0001818435754189944,
+ "loss": 2.2194,
+ "step": 65
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018156424581005588,
+ "loss": 2.1218,
+ "step": 66
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018128491620111731,
+ "loss": 2.2049,
+ "step": 67
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.00018100558659217878,
+ "loss": 2.1521,
+ "step": 68
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.00018072625698324025,
+ "loss": 2.112,
+ "step": 69
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0001804469273743017,
+ "loss": 2.1906,
+ "step": 70
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.00018016759776536313,
+ "loss": 2.1717,
+ "step": 71
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0001798882681564246,
+ "loss": 2.0712,
+ "step": 72
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.00017960893854748603,
+ "loss": 2.141,
+ "step": 73
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0001793296089385475,
+ "loss": 2.0656,
+ "step": 74
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017905027932960894,
+ "loss": 2.1125,
+ "step": 75
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017877094972067038,
+ "loss": 2.0869,
+ "step": 76
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.00017849162011173187,
+ "loss": 2.2478,
+ "step": 77
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0001782122905027933,
+ "loss": 2.1535,
+ "step": 78
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.00017793296089385475,
+ "loss": 2.1927,
+ "step": 79
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017765363128491622,
+ "loss": 2.1213,
+ "step": 80
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017737430167597766,
+ "loss": 2.0981,
+ "step": 81
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017709497206703912,
+ "loss": 2.1828,
+ "step": 82
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017681564245810056,
+ "loss": 2.0562,
+ "step": 83
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.000176536312849162,
+ "loss": 2.1334,
+ "step": 84
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.00017625698324022347,
+ "loss": 2.1225,
+ "step": 85
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.00017597765363128493,
+ "loss": 2.2098,
+ "step": 86
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017569832402234637,
+ "loss": 2.1519,
+ "step": 87
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017541899441340784,
+ "loss": 2.1132,
+ "step": 88
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017513966480446928,
+ "loss": 2.0333,
+ "step": 89
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017486033519553075,
+ "loss": 2.2764,
+ "step": 90
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017458100558659218,
+ "loss": 2.1838,
+ "step": 91
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017430167597765362,
+ "loss": 2.1386,
+ "step": 92
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0001740223463687151,
+ "loss": 2.1034,
+ "step": 93
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.00017374301675977656,
+ "loss": 2.0346,
+ "step": 94
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.000173463687150838,
+ "loss": 2.0274,
+ "step": 95
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.00017318435754189946,
+ "loss": 2.1036,
+ "step": 96
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0001729050279329609,
+ "loss": 2.1208,
+ "step": 97
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.00017262569832402237,
+ "loss": 2.0572,
+ "step": 98
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0001723463687150838,
+ "loss": 2.1702,
+ "step": 99
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.00017206703910614525,
+ "loss": 2.1302,
+ "step": 100
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0001717877094972067,
+ "loss": 2.0175,
+ "step": 101
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.00017150837988826815,
+ "loss": 2.1006,
+ "step": 102
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017122905027932962,
+ "loss": 2.0662,
+ "step": 103
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017094972067039109,
+ "loss": 1.988,
+ "step": 104
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017067039106145253,
+ "loss": 2.1008,
+ "step": 105
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017039106145251396,
+ "loss": 2.1482,
+ "step": 106
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00017011173184357543,
+ "loss": 2.1052,
+ "step": 107
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00016983240223463687,
+ "loss": 2.0978,
+ "step": 108
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016955307262569834,
+ "loss": 2.1303,
+ "step": 109
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016927374301675978,
+ "loss": 2.0794,
+ "step": 110
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.00016899441340782122,
+ "loss": 2.1059,
+ "step": 111
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0001687150837988827,
+ "loss": 1.9642,
+ "step": 112
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.00016843575418994415,
+ "loss": 2.0415,
+ "step": 113
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0001681564245810056,
+ "loss": 2.0795,
+ "step": 114
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.00016787709497206705,
+ "loss": 2.0238,
+ "step": 115
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0001675977653631285,
+ "loss": 2.078,
+ "step": 116
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.00016731843575418996,
+ "loss": 2.1362,
+ "step": 117
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0001670391061452514,
+ "loss": 2.0552,
+ "step": 118
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.00016675977653631284,
+ "loss": 2.213,
+ "step": 119
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0001664804469273743,
+ "loss": 1.9999,
+ "step": 120
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.00016620111731843577,
+ "loss": 2.06,
+ "step": 121
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0001659217877094972,
+ "loss": 2.0177,
+ "step": 122
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016564245810055868,
+ "loss": 2.0504,
+ "step": 123
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016536312849162012,
+ "loss": 2.0585,
+ "step": 124
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016508379888268158,
+ "loss": 2.0273,
+ "step": 125
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016480446927374302,
+ "loss": 2.0549,
+ "step": 126
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.00016452513966480446,
+ "loss": 2.044,
+ "step": 127
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016424581005586593,
+ "loss": 2.0731,
+ "step": 128
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016396648044692737,
+ "loss": 2.0568,
+ "step": 129
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.00016368715083798883,
+ "loss": 2.007,
+ "step": 130
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0001634078212290503,
+ "loss": 2.0707,
+ "step": 131
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.00016312849162011174,
+ "loss": 1.9793,
+ "step": 132
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0001628491620111732,
+ "loss": 2.1311,
+ "step": 133
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016256983240223465,
+ "loss": 2.0016,
+ "step": 134
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016229050279329609,
+ "loss": 1.9945,
+ "step": 135
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.00016201117318435755,
+ "loss": 2.0186,
+ "step": 136
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.000161731843575419,
+ "loss": 2.0971,
+ "step": 137
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.00016145251396648046,
+ "loss": 2.0883,
+ "step": 138
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016117318435754192,
+ "loss": 2.0803,
+ "step": 139
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016089385474860336,
+ "loss": 2.0617,
+ "step": 140
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016061452513966483,
+ "loss": 2.1265,
+ "step": 141
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016033519553072627,
+ "loss": 2.0151,
+ "step": 142
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0001600558659217877,
+ "loss": 1.996,
+ "step": 143
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.00015977653631284918,
+ "loss": 2.0164,
+ "step": 144
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.00015949720670391061,
+ "loss": 2.0314,
+ "step": 145
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015921787709497208,
+ "loss": 1.9501,
+ "step": 146
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015893854748603352,
+ "loss": 2.087,
+ "step": 147
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.000158659217877095,
+ "loss": 2.0262,
+ "step": 148
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.00015837988826815643,
+ "loss": 2.0765,
+ "step": 149
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0001581005586592179,
+ "loss": 2.105,
+ "step": 150
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.00015782122905027933,
+ "loss": 1.9863,
+ "step": 151
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0001575418994413408,
+ "loss": 1.9873,
+ "step": 152
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.00015726256983240224,
+ "loss": 2.0094,
+ "step": 153
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.00015698324022346368,
+ "loss": 1.9141,
+ "step": 154
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015670391061452514,
+ "loss": 1.917,
+ "step": 155
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015642458100558658,
+ "loss": 2.109,
+ "step": 156
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015614525139664805,
+ "loss": 1.9799,
+ "step": 157
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015586592178770952,
+ "loss": 1.9571,
+ "step": 158
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015558659217877096,
+ "loss": 1.9931,
+ "step": 159
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015530726256983242,
+ "loss": 2.1004,
+ "step": 160
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.00015502793296089386,
+ "loss": 2.0385,
+ "step": 161
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0001547486033519553,
+ "loss": 1.9751,
+ "step": 162
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.00015446927374301677,
+ "loss": 2.0544,
+ "step": 163
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0001541899441340782,
+ "loss": 2.0069,
+ "step": 164
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.00015391061452513967,
+ "loss": 1.9576,
+ "step": 165
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015363128491620114,
+ "loss": 1.8991,
+ "step": 166
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015335195530726258,
+ "loss": 1.9336,
+ "step": 167
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015307262569832405,
+ "loss": 1.9736,
+ "step": 168
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015279329608938548,
+ "loss": 1.9702,
+ "step": 169
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.00015251396648044692,
+ "loss": 1.9055,
+ "step": 170
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0001522346368715084,
+ "loss": 2.0503,
+ "step": 171
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.00015195530726256983,
+ "loss": 2.0039,
+ "step": 172
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0001516759776536313,
+ "loss": 1.9406,
+ "step": 173
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.00015139664804469274,
+ "loss": 2.0525,
+ "step": 174
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0001511173184357542,
+ "loss": 1.9234,
+ "step": 175
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.00015083798882681567,
+ "loss": 1.8614,
+ "step": 176
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0001505586592178771,
+ "loss": 1.9616,
+ "step": 177
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.00015027932960893855,
+ "loss": 1.9509,
+ "step": 178
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.9592,
+ "step": 179
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014972067039106145,
+ "loss": 1.8991,
+ "step": 180
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014944134078212292,
+ "loss": 1.9127,
+ "step": 181
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.00014916201117318436,
+ "loss": 1.8982,
+ "step": 182
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0001488826815642458,
+ "loss": 1.9534,
+ "step": 183
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0001486033519553073,
+ "loss": 1.7794,
+ "step": 184
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.00014832402234636873,
+ "loss": 1.7958,
+ "step": 185
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014804469273743017,
+ "loss": 1.8282,
+ "step": 186
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014776536312849164,
+ "loss": 2.0423,
+ "step": 187
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.00014748603351955308,
+ "loss": 1.9282,
+ "step": 188
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014720670391061454,
+ "loss": 1.9072,
+ "step": 189
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014692737430167598,
+ "loss": 1.8665,
+ "step": 190
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.00014664804469273742,
+ "loss": 1.9021,
+ "step": 191
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.0001463687150837989,
+ "loss": 1.7308,
+ "step": 192
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.00014608938547486035,
+ "loss": 1.9165,
+ "step": 193
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.0001458100558659218,
+ "loss": 1.842,
+ "step": 194
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.00014553072625698326,
+ "loss": 1.9128,
+ "step": 195
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.0001452513966480447,
+ "loss": 1.8005,
+ "step": 196
+ },
+ {
+ "epoch": 1.1,
+ "learning_rate": 0.00014497206703910614,
+ "loss": 1.8547,
+ "step": 197
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.0001446927374301676,
+ "loss": 1.9042,
+ "step": 198
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.00014441340782122905,
+ "loss": 1.8609,
+ "step": 199
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.0001441340782122905,
+ "loss": 1.9591,
+ "step": 200
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 716,
+ "num_train_epochs": 4,
+ "save_steps": 100,
+ "total_flos": 1.0290223796256768e+17,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-200/training_args.bin b/checkpoint-200/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8672c716e925d0028b4938db147703f58656ff7
--- /dev/null
+++ b/checkpoint-200/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:188ae1c421cc0c6435d1f71d8d3423ac4abc7dba0e6fc2efcbc4dbe77c741317
+size 4027
diff --git a/checkpoint-300/README.md b/checkpoint-300/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..2f257a448caef1c59022426ad3ec9ba80acb3821
--- /dev/null
+++ b/checkpoint-300/README.md
@@ -0,0 +1,60 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+### Framework versions
+
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+
+- PEFT 0.6.0.dev0
diff --git a/checkpoint-300/adapter_config.json b/checkpoint-300/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c661f74e40394c079de68f07861d73c341803b67
--- /dev/null
+++ b/checkpoint-300/adapter_config.json
@@ -0,0 +1,25 @@
+{
+ "alpha_pattern": {},
+ "auto_mapping": null,
+ "base_model_name_or_path": "bigscience/bloomz-3b",
+ "bias": "none",
+ "fan_in_fan_out": false,
+ "inference_mode": true,
+ "init_lora_weights": true,
+ "layers_pattern": null,
+ "layers_to_transform": null,
+ "lora_alpha": 16,
+ "lora_dropout": 0.0,
+ "modules_to_save": null,
+ "peft_type": "LORA",
+ "r": 8,
+ "rank_pattern": {},
+ "revision": null,
+ "target_modules": [
+ "dense_4h_to_h",
+ "dense",
+ "dense_h_to_4h",
+ "query_key_value"
+ ],
+ "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/checkpoint-300/adapter_model.bin b/checkpoint-300/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..f32eee84dcfc362bb3dd9f9d9127291a2f42668f
--- /dev/null
+++ b/checkpoint-300/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b39b2ac3c3f75772f547048fda7ecb323076b9689db7d79915bba156dc508f2f
+size 39409357
diff --git a/checkpoint-300/optimizer.pt b/checkpoint-300/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..3d79ee6144baeefb728bb603d5ca209014a1847b
--- /dev/null
+++ b/checkpoint-300/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c734ecaa394370ee4bcd94cc0b2ae016a26765122f3f76327b28c23f96a22732
+size 78844421
diff --git a/checkpoint-300/rng_state.pth b/checkpoint-300/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b9111227e3b79b9bc3e2a642832c3e49e36216d4
--- /dev/null
+++ b/checkpoint-300/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:972139d83957a9cf2600cb6eeca17287d7a5377c33a53500ae7e13fe830ad36b
+size 14575
diff --git a/checkpoint-300/scheduler.pt b/checkpoint-300/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..ba7790505ccbdccc3664cac72a2743abf3f54ee6
--- /dev/null
+++ b/checkpoint-300/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d8dcaf05375bb59f736a94e8f8b03d33cdc87bc02411e6527a29996e0a68b3b
+size 627
diff --git a/checkpoint-300/special_tokens_map.json b/checkpoint-300/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdafe480f024ff444c7492147536765ce5d55a2d
--- /dev/null
+++ b/checkpoint-300/special_tokens_map.json
@@ -0,0 +1,6 @@
+{
+ "bos_token": "",
+ "eos_token": "",
+ "pad_token": "",
+ "unk_token": ""
+}
diff --git a/checkpoint-300/tokenizer.json b/checkpoint-300/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..673c31abdeadf6576c3c754df86459e1ad64e207
--- /dev/null
+++ b/checkpoint-300/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85b00d7db4df5df2e3f01cacc3feda246002a672f3356eec7f4b04a22eb0dfbe
+size 14500570
diff --git a/checkpoint-300/tokenizer_config.json b/checkpoint-300/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b56cc9c2965c07132c35df3e2972e93d98c82c3
--- /dev/null
+++ b/checkpoint-300/tokenizer_config.json
@@ -0,0 +1,10 @@
+{
+ "add_prefix_space": false,
+ "bos_token": "",
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "",
+ "model_max_length": 1000000000000000019884624838656,
+ "pad_token": "",
+ "tokenizer_class": "BloomTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoint-300/trainer_state.json b/checkpoint-300/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf5790647f2fda83e639ffc04ccde4b8937523ee
--- /dev/null
+++ b/checkpoint-300/trainer_state.json
@@ -0,0 +1,1819 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 1.675977653631285,
+ "eval_steps": 500,
+ "global_step": 300,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019972067039106145,
+ "loss": 2.6443,
+ "step": 1
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019944134078212292,
+ "loss": 2.4104,
+ "step": 2
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019916201117318435,
+ "loss": 2.4975,
+ "step": 3
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019888268156424582,
+ "loss": 2.3513,
+ "step": 4
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0001986033519553073,
+ "loss": 2.4274,
+ "step": 5
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.00019832402234636873,
+ "loss": 2.3628,
+ "step": 6
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0001980446927374302,
+ "loss": 2.3567,
+ "step": 7
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.00019776536312849163,
+ "loss": 2.4121,
+ "step": 8
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.00019748603351955307,
+ "loss": 2.4033,
+ "step": 9
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019720670391061454,
+ "loss": 2.2805,
+ "step": 10
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019692737430167598,
+ "loss": 2.2639,
+ "step": 11
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019664804469273744,
+ "loss": 2.2724,
+ "step": 12
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019636871508379888,
+ "loss": 2.332,
+ "step": 13
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019608938547486035,
+ "loss": 2.2261,
+ "step": 14
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019581005586592182,
+ "loss": 2.2208,
+ "step": 15
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.00019553072625698326,
+ "loss": 2.3351,
+ "step": 16
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0001952513966480447,
+ "loss": 2.2475,
+ "step": 17
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.00019497206703910616,
+ "loss": 2.3283,
+ "step": 18
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0001946927374301676,
+ "loss": 2.1346,
+ "step": 19
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.00019441340782122907,
+ "loss": 2.131,
+ "step": 20
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0001941340782122905,
+ "loss": 2.1718,
+ "step": 21
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.00019385474860335195,
+ "loss": 2.2446,
+ "step": 22
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0001935754189944134,
+ "loss": 2.306,
+ "step": 23
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.00019329608938547488,
+ "loss": 2.1908,
+ "step": 24
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.00019301675977653632,
+ "loss": 2.2844,
+ "step": 25
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019273743016759779,
+ "loss": 2.2235,
+ "step": 26
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019245810055865922,
+ "loss": 2.1842,
+ "step": 27
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019217877094972066,
+ "loss": 2.2675,
+ "step": 28
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019189944134078213,
+ "loss": 2.2532,
+ "step": 29
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019162011173184357,
+ "loss": 2.1788,
+ "step": 30
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019134078212290504,
+ "loss": 2.2494,
+ "step": 31
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0001910614525139665,
+ "loss": 2.1995,
+ "step": 32
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.00019078212290502794,
+ "loss": 2.1451,
+ "step": 33
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0001905027932960894,
+ "loss": 2.223,
+ "step": 34
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.00019022346368715085,
+ "loss": 2.2854,
+ "step": 35
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0001899441340782123,
+ "loss": 2.2265,
+ "step": 36
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.00018966480446927375,
+ "loss": 2.1214,
+ "step": 37
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0001893854748603352,
+ "loss": 2.1898,
+ "step": 38
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.00018910614525139666,
+ "loss": 2.1974,
+ "step": 39
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0001888268156424581,
+ "loss": 2.2259,
+ "step": 40
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018854748603351957,
+ "loss": 2.2094,
+ "step": 41
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018826815642458103,
+ "loss": 2.1731,
+ "step": 42
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.00018798882681564247,
+ "loss": 2.2373,
+ "step": 43
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0001877094972067039,
+ "loss": 2.2295,
+ "step": 44
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.00018743016759776538,
+ "loss": 2.1947,
+ "step": 45
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018715083798882682,
+ "loss": 2.2115,
+ "step": 46
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018687150837988828,
+ "loss": 2.1224,
+ "step": 47
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018659217877094972,
+ "loss": 2.2137,
+ "step": 48
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018631284916201116,
+ "loss": 2.2338,
+ "step": 49
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.00018603351955307266,
+ "loss": 2.1298,
+ "step": 50
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0001857541899441341,
+ "loss": 2.0883,
+ "step": 51
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.00018547486033519553,
+ "loss": 2.1216,
+ "step": 52
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.000185195530726257,
+ "loss": 2.2112,
+ "step": 53
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.00018491620111731844,
+ "loss": 2.1224,
+ "step": 54
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0001846368715083799,
+ "loss": 2.2375,
+ "step": 55
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.00018435754189944135,
+ "loss": 2.2235,
+ "step": 56
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018407821229050279,
+ "loss": 2.1682,
+ "step": 57
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018379888268156425,
+ "loss": 2.2077,
+ "step": 58
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.00018351955307262572,
+ "loss": 2.1596,
+ "step": 59
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018324022346368716,
+ "loss": 2.1311,
+ "step": 60
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018296089385474862,
+ "loss": 2.1333,
+ "step": 61
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018268156424581006,
+ "loss": 2.0901,
+ "step": 62
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018240223463687153,
+ "loss": 2.1971,
+ "step": 63
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.00018212290502793297,
+ "loss": 2.2602,
+ "step": 64
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0001818435754189944,
+ "loss": 2.2194,
+ "step": 65
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018156424581005588,
+ "loss": 2.1218,
+ "step": 66
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018128491620111731,
+ "loss": 2.2049,
+ "step": 67
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.00018100558659217878,
+ "loss": 2.1521,
+ "step": 68
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.00018072625698324025,
+ "loss": 2.112,
+ "step": 69
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0001804469273743017,
+ "loss": 2.1906,
+ "step": 70
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.00018016759776536313,
+ "loss": 2.1717,
+ "step": 71
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0001798882681564246,
+ "loss": 2.0712,
+ "step": 72
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.00017960893854748603,
+ "loss": 2.141,
+ "step": 73
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0001793296089385475,
+ "loss": 2.0656,
+ "step": 74
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017905027932960894,
+ "loss": 2.1125,
+ "step": 75
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017877094972067038,
+ "loss": 2.0869,
+ "step": 76
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.00017849162011173187,
+ "loss": 2.2478,
+ "step": 77
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0001782122905027933,
+ "loss": 2.1535,
+ "step": 78
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.00017793296089385475,
+ "loss": 2.1927,
+ "step": 79
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017765363128491622,
+ "loss": 2.1213,
+ "step": 80
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017737430167597766,
+ "loss": 2.0981,
+ "step": 81
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017709497206703912,
+ "loss": 2.1828,
+ "step": 82
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017681564245810056,
+ "loss": 2.0562,
+ "step": 83
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.000176536312849162,
+ "loss": 2.1334,
+ "step": 84
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.00017625698324022347,
+ "loss": 2.1225,
+ "step": 85
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.00017597765363128493,
+ "loss": 2.2098,
+ "step": 86
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017569832402234637,
+ "loss": 2.1519,
+ "step": 87
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017541899441340784,
+ "loss": 2.1132,
+ "step": 88
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017513966480446928,
+ "loss": 2.0333,
+ "step": 89
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017486033519553075,
+ "loss": 2.2764,
+ "step": 90
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017458100558659218,
+ "loss": 2.1838,
+ "step": 91
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017430167597765362,
+ "loss": 2.1386,
+ "step": 92
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0001740223463687151,
+ "loss": 2.1034,
+ "step": 93
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.00017374301675977656,
+ "loss": 2.0346,
+ "step": 94
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.000173463687150838,
+ "loss": 2.0274,
+ "step": 95
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.00017318435754189946,
+ "loss": 2.1036,
+ "step": 96
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0001729050279329609,
+ "loss": 2.1208,
+ "step": 97
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.00017262569832402237,
+ "loss": 2.0572,
+ "step": 98
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0001723463687150838,
+ "loss": 2.1702,
+ "step": 99
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.00017206703910614525,
+ "loss": 2.1302,
+ "step": 100
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0001717877094972067,
+ "loss": 2.0175,
+ "step": 101
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.00017150837988826815,
+ "loss": 2.1006,
+ "step": 102
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017122905027932962,
+ "loss": 2.0662,
+ "step": 103
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017094972067039109,
+ "loss": 1.988,
+ "step": 104
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017067039106145253,
+ "loss": 2.1008,
+ "step": 105
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017039106145251396,
+ "loss": 2.1482,
+ "step": 106
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00017011173184357543,
+ "loss": 2.1052,
+ "step": 107
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00016983240223463687,
+ "loss": 2.0978,
+ "step": 108
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016955307262569834,
+ "loss": 2.1303,
+ "step": 109
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016927374301675978,
+ "loss": 2.0794,
+ "step": 110
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.00016899441340782122,
+ "loss": 2.1059,
+ "step": 111
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0001687150837988827,
+ "loss": 1.9642,
+ "step": 112
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.00016843575418994415,
+ "loss": 2.0415,
+ "step": 113
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0001681564245810056,
+ "loss": 2.0795,
+ "step": 114
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.00016787709497206705,
+ "loss": 2.0238,
+ "step": 115
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0001675977653631285,
+ "loss": 2.078,
+ "step": 116
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.00016731843575418996,
+ "loss": 2.1362,
+ "step": 117
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0001670391061452514,
+ "loss": 2.0552,
+ "step": 118
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.00016675977653631284,
+ "loss": 2.213,
+ "step": 119
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0001664804469273743,
+ "loss": 1.9999,
+ "step": 120
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.00016620111731843577,
+ "loss": 2.06,
+ "step": 121
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0001659217877094972,
+ "loss": 2.0177,
+ "step": 122
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016564245810055868,
+ "loss": 2.0504,
+ "step": 123
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016536312849162012,
+ "loss": 2.0585,
+ "step": 124
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016508379888268158,
+ "loss": 2.0273,
+ "step": 125
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016480446927374302,
+ "loss": 2.0549,
+ "step": 126
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.00016452513966480446,
+ "loss": 2.044,
+ "step": 127
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016424581005586593,
+ "loss": 2.0731,
+ "step": 128
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016396648044692737,
+ "loss": 2.0568,
+ "step": 129
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.00016368715083798883,
+ "loss": 2.007,
+ "step": 130
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0001634078212290503,
+ "loss": 2.0707,
+ "step": 131
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.00016312849162011174,
+ "loss": 1.9793,
+ "step": 132
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0001628491620111732,
+ "loss": 2.1311,
+ "step": 133
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016256983240223465,
+ "loss": 2.0016,
+ "step": 134
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016229050279329609,
+ "loss": 1.9945,
+ "step": 135
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.00016201117318435755,
+ "loss": 2.0186,
+ "step": 136
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.000161731843575419,
+ "loss": 2.0971,
+ "step": 137
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.00016145251396648046,
+ "loss": 2.0883,
+ "step": 138
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016117318435754192,
+ "loss": 2.0803,
+ "step": 139
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016089385474860336,
+ "loss": 2.0617,
+ "step": 140
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016061452513966483,
+ "loss": 2.1265,
+ "step": 141
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016033519553072627,
+ "loss": 2.0151,
+ "step": 142
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0001600558659217877,
+ "loss": 1.996,
+ "step": 143
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.00015977653631284918,
+ "loss": 2.0164,
+ "step": 144
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.00015949720670391061,
+ "loss": 2.0314,
+ "step": 145
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015921787709497208,
+ "loss": 1.9501,
+ "step": 146
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015893854748603352,
+ "loss": 2.087,
+ "step": 147
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.000158659217877095,
+ "loss": 2.0262,
+ "step": 148
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.00015837988826815643,
+ "loss": 2.0765,
+ "step": 149
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0001581005586592179,
+ "loss": 2.105,
+ "step": 150
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.00015782122905027933,
+ "loss": 1.9863,
+ "step": 151
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0001575418994413408,
+ "loss": 1.9873,
+ "step": 152
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.00015726256983240224,
+ "loss": 2.0094,
+ "step": 153
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.00015698324022346368,
+ "loss": 1.9141,
+ "step": 154
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015670391061452514,
+ "loss": 1.917,
+ "step": 155
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015642458100558658,
+ "loss": 2.109,
+ "step": 156
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015614525139664805,
+ "loss": 1.9799,
+ "step": 157
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015586592178770952,
+ "loss": 1.9571,
+ "step": 158
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015558659217877096,
+ "loss": 1.9931,
+ "step": 159
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015530726256983242,
+ "loss": 2.1004,
+ "step": 160
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.00015502793296089386,
+ "loss": 2.0385,
+ "step": 161
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0001547486033519553,
+ "loss": 1.9751,
+ "step": 162
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.00015446927374301677,
+ "loss": 2.0544,
+ "step": 163
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0001541899441340782,
+ "loss": 2.0069,
+ "step": 164
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.00015391061452513967,
+ "loss": 1.9576,
+ "step": 165
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015363128491620114,
+ "loss": 1.8991,
+ "step": 166
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015335195530726258,
+ "loss": 1.9336,
+ "step": 167
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015307262569832405,
+ "loss": 1.9736,
+ "step": 168
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015279329608938548,
+ "loss": 1.9702,
+ "step": 169
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.00015251396648044692,
+ "loss": 1.9055,
+ "step": 170
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0001522346368715084,
+ "loss": 2.0503,
+ "step": 171
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.00015195530726256983,
+ "loss": 2.0039,
+ "step": 172
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0001516759776536313,
+ "loss": 1.9406,
+ "step": 173
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.00015139664804469274,
+ "loss": 2.0525,
+ "step": 174
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0001511173184357542,
+ "loss": 1.9234,
+ "step": 175
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.00015083798882681567,
+ "loss": 1.8614,
+ "step": 176
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0001505586592178771,
+ "loss": 1.9616,
+ "step": 177
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.00015027932960893855,
+ "loss": 1.9509,
+ "step": 178
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.9592,
+ "step": 179
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014972067039106145,
+ "loss": 1.8991,
+ "step": 180
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014944134078212292,
+ "loss": 1.9127,
+ "step": 181
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.00014916201117318436,
+ "loss": 1.8982,
+ "step": 182
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0001488826815642458,
+ "loss": 1.9534,
+ "step": 183
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0001486033519553073,
+ "loss": 1.7794,
+ "step": 184
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.00014832402234636873,
+ "loss": 1.7958,
+ "step": 185
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014804469273743017,
+ "loss": 1.8282,
+ "step": 186
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014776536312849164,
+ "loss": 2.0423,
+ "step": 187
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.00014748603351955308,
+ "loss": 1.9282,
+ "step": 188
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014720670391061454,
+ "loss": 1.9072,
+ "step": 189
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014692737430167598,
+ "loss": 1.8665,
+ "step": 190
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.00014664804469273742,
+ "loss": 1.9021,
+ "step": 191
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.0001463687150837989,
+ "loss": 1.7308,
+ "step": 192
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.00014608938547486035,
+ "loss": 1.9165,
+ "step": 193
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.0001458100558659218,
+ "loss": 1.842,
+ "step": 194
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.00014553072625698326,
+ "loss": 1.9128,
+ "step": 195
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.0001452513966480447,
+ "loss": 1.8005,
+ "step": 196
+ },
+ {
+ "epoch": 1.1,
+ "learning_rate": 0.00014497206703910614,
+ "loss": 1.8547,
+ "step": 197
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.0001446927374301676,
+ "loss": 1.9042,
+ "step": 198
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.00014441340782122905,
+ "loss": 1.8609,
+ "step": 199
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.0001441340782122905,
+ "loss": 1.9591,
+ "step": 200
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.00014385474860335195,
+ "loss": 1.8722,
+ "step": 201
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014357541899441342,
+ "loss": 1.8535,
+ "step": 202
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014329608938547488,
+ "loss": 1.8676,
+ "step": 203
+ },
+ {
+ "epoch": 1.14,
+ "learning_rate": 0.00014301675977653632,
+ "loss": 1.8976,
+ "step": 204
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014273743016759776,
+ "loss": 1.7723,
+ "step": 205
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014245810055865923,
+ "loss": 1.8165,
+ "step": 206
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014217877094972067,
+ "loss": 1.7811,
+ "step": 207
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014189944134078214,
+ "loss": 1.908,
+ "step": 208
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014162011173184357,
+ "loss": 1.7663,
+ "step": 209
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014134078212290501,
+ "loss": 1.6779,
+ "step": 210
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.0001410614525139665,
+ "loss": 1.9039,
+ "step": 211
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.00014078212290502795,
+ "loss": 1.8033,
+ "step": 212
+ },
+ {
+ "epoch": 1.19,
+ "learning_rate": 0.00014050279329608939,
+ "loss": 1.8251,
+ "step": 213
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.00014022346368715085,
+ "loss": 1.8505,
+ "step": 214
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.0001399441340782123,
+ "loss": 1.8147,
+ "step": 215
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.00013966480446927376,
+ "loss": 1.8419,
+ "step": 216
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.0001393854748603352,
+ "loss": 1.8401,
+ "step": 217
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.00013910614525139664,
+ "loss": 1.8912,
+ "step": 218
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.0001388268156424581,
+ "loss": 1.7548,
+ "step": 219
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.00013854748603351957,
+ "loss": 1.8741,
+ "step": 220
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.000138268156424581,
+ "loss": 1.9549,
+ "step": 221
+ },
+ {
+ "epoch": 1.24,
+ "learning_rate": 0.00013798882681564248,
+ "loss": 1.9093,
+ "step": 222
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013770949720670392,
+ "loss": 1.7896,
+ "step": 223
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013743016759776538,
+ "loss": 1.8491,
+ "step": 224
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013715083798882682,
+ "loss": 1.7851,
+ "step": 225
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013687150837988826,
+ "loss": 1.6992,
+ "step": 226
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013659217877094973,
+ "loss": 1.9765,
+ "step": 227
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013631284916201117,
+ "loss": 1.8179,
+ "step": 228
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.00013603351955307263,
+ "loss": 1.8548,
+ "step": 229
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.0001357541899441341,
+ "loss": 1.8843,
+ "step": 230
+ },
+ {
+ "epoch": 1.29,
+ "learning_rate": 0.00013547486033519554,
+ "loss": 1.9105,
+ "step": 231
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013519553072625698,
+ "loss": 1.8748,
+ "step": 232
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013491620111731844,
+ "loss": 1.7976,
+ "step": 233
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013463687150837988,
+ "loss": 1.7369,
+ "step": 234
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013435754189944135,
+ "loss": 1.7808,
+ "step": 235
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.0001340782122905028,
+ "loss": 1.8385,
+ "step": 236
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.00013379888268156423,
+ "loss": 1.8295,
+ "step": 237
+ },
+ {
+ "epoch": 1.33,
+ "learning_rate": 0.00013351955307262572,
+ "loss": 1.757,
+ "step": 238
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.00013324022346368716,
+ "loss": 1.7904,
+ "step": 239
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.0001329608938547486,
+ "loss": 1.7632,
+ "step": 240
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.00013268156424581007,
+ "loss": 1.7867,
+ "step": 241
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.0001324022346368715,
+ "loss": 1.8259,
+ "step": 242
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.00013212290502793297,
+ "loss": 1.6655,
+ "step": 243
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.0001318435754189944,
+ "loss": 1.6848,
+ "step": 244
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013156424581005585,
+ "loss": 1.7931,
+ "step": 245
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013128491620111732,
+ "loss": 1.7868,
+ "step": 246
+ },
+ {
+ "epoch": 1.38,
+ "learning_rate": 0.00013100558659217879,
+ "loss": 1.7732,
+ "step": 247
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.00013072625698324022,
+ "loss": 1.7851,
+ "step": 248
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.0001304469273743017,
+ "loss": 1.7406,
+ "step": 249
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.00013016759776536313,
+ "loss": 1.5853,
+ "step": 250
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.0001298882681564246,
+ "loss": 1.8271,
+ "step": 251
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012960893854748604,
+ "loss": 1.6054,
+ "step": 252
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012932960893854748,
+ "loss": 1.6884,
+ "step": 253
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012905027932960894,
+ "loss": 1.7333,
+ "step": 254
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012877094972067038,
+ "loss": 1.803,
+ "step": 255
+ },
+ {
+ "epoch": 1.43,
+ "learning_rate": 0.00012849162011173185,
+ "loss": 1.63,
+ "step": 256
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012821229050279331,
+ "loss": 1.8023,
+ "step": 257
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012793296089385475,
+ "loss": 1.762,
+ "step": 258
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012765363128491622,
+ "loss": 1.7565,
+ "step": 259
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012737430167597766,
+ "loss": 1.5937,
+ "step": 260
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.0001270949720670391,
+ "loss": 1.7208,
+ "step": 261
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.00012681564245810057,
+ "loss": 1.8097,
+ "step": 262
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.000126536312849162,
+ "loss": 1.7042,
+ "step": 263
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.00012625698324022347,
+ "loss": 1.7892,
+ "step": 264
+ },
+ {
+ "epoch": 1.48,
+ "learning_rate": 0.00012597765363128494,
+ "loss": 1.752,
+ "step": 265
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012569832402234638,
+ "loss": 1.7333,
+ "step": 266
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012541899441340784,
+ "loss": 1.7492,
+ "step": 267
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012513966480446928,
+ "loss": 1.818,
+ "step": 268
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012486033519553072,
+ "loss": 1.7146,
+ "step": 269
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.0001245810055865922,
+ "loss": 1.7958,
+ "step": 270
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.00012430167597765363,
+ "loss": 1.7212,
+ "step": 271
+ },
+ {
+ "epoch": 1.52,
+ "learning_rate": 0.0001240223463687151,
+ "loss": 1.8835,
+ "step": 272
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.00012374301675977656,
+ "loss": 1.7486,
+ "step": 273
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.000123463687150838,
+ "loss": 1.8103,
+ "step": 274
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.00012318435754189944,
+ "loss": 1.7557,
+ "step": 275
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.0001229050279329609,
+ "loss": 1.6198,
+ "step": 276
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.00012262569832402235,
+ "loss": 1.6971,
+ "step": 277
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.0001223463687150838,
+ "loss": 1.668,
+ "step": 278
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.00012206703910614525,
+ "loss": 1.8795,
+ "step": 279
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.0001217877094972067,
+ "loss": 1.6412,
+ "step": 280
+ },
+ {
+ "epoch": 1.57,
+ "learning_rate": 0.00012150837988826816,
+ "loss": 1.7497,
+ "step": 281
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012122905027932962,
+ "loss": 1.5577,
+ "step": 282
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012094972067039108,
+ "loss": 1.8049,
+ "step": 283
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012067039106145253,
+ "loss": 1.6834,
+ "step": 284
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012039106145251397,
+ "loss": 1.7978,
+ "step": 285
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00012011173184357542,
+ "loss": 1.6558,
+ "step": 286
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00011983240223463687,
+ "loss": 1.7561,
+ "step": 287
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011955307262569833,
+ "loss": 1.6807,
+ "step": 288
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011927374301675978,
+ "loss": 1.6937,
+ "step": 289
+ },
+ {
+ "epoch": 1.62,
+ "learning_rate": 0.00011899441340782122,
+ "loss": 1.6591,
+ "step": 290
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.0001187150837988827,
+ "loss": 1.6771,
+ "step": 291
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.00011843575418994415,
+ "loss": 1.7743,
+ "step": 292
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011815642458100559,
+ "loss": 1.5857,
+ "step": 293
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011787709497206705,
+ "loss": 1.6999,
+ "step": 294
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.0001175977653631285,
+ "loss": 1.5661,
+ "step": 295
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.00011731843575418995,
+ "loss": 1.7235,
+ "step": 296
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.0001170391061452514,
+ "loss": 1.607,
+ "step": 297
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.00011675977653631284,
+ "loss": 1.68,
+ "step": 298
+ },
+ {
+ "epoch": 1.67,
+ "learning_rate": 0.0001164804469273743,
+ "loss": 1.6938,
+ "step": 299
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011620111731843578,
+ "loss": 1.6315,
+ "step": 300
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 716,
+ "num_train_epochs": 4,
+ "save_steps": 100,
+ "total_flos": 1.5369996759656448e+17,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-300/training_args.bin b/checkpoint-300/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8672c716e925d0028b4938db147703f58656ff7
--- /dev/null
+++ b/checkpoint-300/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:188ae1c421cc0c6435d1f71d8d3423ac4abc7dba0e6fc2efcbc4dbe77c741317
+size 4027
diff --git a/checkpoint-400/README.md b/checkpoint-400/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..2f257a448caef1c59022426ad3ec9ba80acb3821
--- /dev/null
+++ b/checkpoint-400/README.md
@@ -0,0 +1,60 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+### Framework versions
+
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+
+- PEFT 0.6.0.dev0
diff --git a/checkpoint-400/adapter_config.json b/checkpoint-400/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c661f74e40394c079de68f07861d73c341803b67
--- /dev/null
+++ b/checkpoint-400/adapter_config.json
@@ -0,0 +1,25 @@
+{
+ "alpha_pattern": {},
+ "auto_mapping": null,
+ "base_model_name_or_path": "bigscience/bloomz-3b",
+ "bias": "none",
+ "fan_in_fan_out": false,
+ "inference_mode": true,
+ "init_lora_weights": true,
+ "layers_pattern": null,
+ "layers_to_transform": null,
+ "lora_alpha": 16,
+ "lora_dropout": 0.0,
+ "modules_to_save": null,
+ "peft_type": "LORA",
+ "r": 8,
+ "rank_pattern": {},
+ "revision": null,
+ "target_modules": [
+ "dense_4h_to_h",
+ "dense",
+ "dense_h_to_4h",
+ "query_key_value"
+ ],
+ "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/checkpoint-400/adapter_model.bin b/checkpoint-400/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..1b6fe766ab6e3bde1e95c1168e8b27f515e22af0
--- /dev/null
+++ b/checkpoint-400/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7b701342ea311c1e25f0f6953505f048df3f56027dc5c764b95df4491bfde1c
+size 39409357
diff --git a/checkpoint-400/optimizer.pt b/checkpoint-400/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..0119015704369e84c41214b2eda49f7eed10bbe8
--- /dev/null
+++ b/checkpoint-400/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da79792a601e6737b980ff24e453353405dce9a807a0964433ac96d5a97b84aa
+size 78844421
diff --git a/checkpoint-400/rng_state.pth b/checkpoint-400/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e0df6592e27bef84a4beb9293e7b49666a4d652e
--- /dev/null
+++ b/checkpoint-400/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7508d4b8dd267de5cc58e972da25236687927651336a28f292c92f7f23951475
+size 14575
diff --git a/checkpoint-400/scheduler.pt b/checkpoint-400/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..b3ccbb3bd3df5add5130e76711d025db7007acb0
--- /dev/null
+++ b/checkpoint-400/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04fdaf305b17644f9d215d28be45b77e6d0f8e4b5adc1c7045c50a61bd8f3c5b
+size 627
diff --git a/checkpoint-400/special_tokens_map.json b/checkpoint-400/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdafe480f024ff444c7492147536765ce5d55a2d
--- /dev/null
+++ b/checkpoint-400/special_tokens_map.json
@@ -0,0 +1,6 @@
+{
+ "bos_token": "",
+ "eos_token": "",
+ "pad_token": "",
+ "unk_token": ""
+}
diff --git a/checkpoint-400/tokenizer.json b/checkpoint-400/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..673c31abdeadf6576c3c754df86459e1ad64e207
--- /dev/null
+++ b/checkpoint-400/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85b00d7db4df5df2e3f01cacc3feda246002a672f3356eec7f4b04a22eb0dfbe
+size 14500570
diff --git a/checkpoint-400/tokenizer_config.json b/checkpoint-400/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b56cc9c2965c07132c35df3e2972e93d98c82c3
--- /dev/null
+++ b/checkpoint-400/tokenizer_config.json
@@ -0,0 +1,10 @@
+{
+ "add_prefix_space": false,
+ "bos_token": "",
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "",
+ "model_max_length": 1000000000000000019884624838656,
+ "pad_token": "",
+ "tokenizer_class": "BloomTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoint-400/trainer_state.json b/checkpoint-400/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..f60e92395bfa70b2808af040d92ac135cc2b0e93
--- /dev/null
+++ b/checkpoint-400/trainer_state.json
@@ -0,0 +1,2419 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 2.2346368715083798,
+ "eval_steps": 500,
+ "global_step": 400,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019972067039106145,
+ "loss": 2.6443,
+ "step": 1
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019944134078212292,
+ "loss": 2.4104,
+ "step": 2
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019916201117318435,
+ "loss": 2.4975,
+ "step": 3
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019888268156424582,
+ "loss": 2.3513,
+ "step": 4
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0001986033519553073,
+ "loss": 2.4274,
+ "step": 5
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.00019832402234636873,
+ "loss": 2.3628,
+ "step": 6
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0001980446927374302,
+ "loss": 2.3567,
+ "step": 7
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.00019776536312849163,
+ "loss": 2.4121,
+ "step": 8
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.00019748603351955307,
+ "loss": 2.4033,
+ "step": 9
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019720670391061454,
+ "loss": 2.2805,
+ "step": 10
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019692737430167598,
+ "loss": 2.2639,
+ "step": 11
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019664804469273744,
+ "loss": 2.2724,
+ "step": 12
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019636871508379888,
+ "loss": 2.332,
+ "step": 13
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019608938547486035,
+ "loss": 2.2261,
+ "step": 14
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019581005586592182,
+ "loss": 2.2208,
+ "step": 15
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.00019553072625698326,
+ "loss": 2.3351,
+ "step": 16
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0001952513966480447,
+ "loss": 2.2475,
+ "step": 17
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.00019497206703910616,
+ "loss": 2.3283,
+ "step": 18
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0001946927374301676,
+ "loss": 2.1346,
+ "step": 19
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.00019441340782122907,
+ "loss": 2.131,
+ "step": 20
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0001941340782122905,
+ "loss": 2.1718,
+ "step": 21
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.00019385474860335195,
+ "loss": 2.2446,
+ "step": 22
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0001935754189944134,
+ "loss": 2.306,
+ "step": 23
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.00019329608938547488,
+ "loss": 2.1908,
+ "step": 24
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.00019301675977653632,
+ "loss": 2.2844,
+ "step": 25
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019273743016759779,
+ "loss": 2.2235,
+ "step": 26
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019245810055865922,
+ "loss": 2.1842,
+ "step": 27
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019217877094972066,
+ "loss": 2.2675,
+ "step": 28
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019189944134078213,
+ "loss": 2.2532,
+ "step": 29
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019162011173184357,
+ "loss": 2.1788,
+ "step": 30
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019134078212290504,
+ "loss": 2.2494,
+ "step": 31
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0001910614525139665,
+ "loss": 2.1995,
+ "step": 32
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.00019078212290502794,
+ "loss": 2.1451,
+ "step": 33
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0001905027932960894,
+ "loss": 2.223,
+ "step": 34
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.00019022346368715085,
+ "loss": 2.2854,
+ "step": 35
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0001899441340782123,
+ "loss": 2.2265,
+ "step": 36
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.00018966480446927375,
+ "loss": 2.1214,
+ "step": 37
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0001893854748603352,
+ "loss": 2.1898,
+ "step": 38
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.00018910614525139666,
+ "loss": 2.1974,
+ "step": 39
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0001888268156424581,
+ "loss": 2.2259,
+ "step": 40
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018854748603351957,
+ "loss": 2.2094,
+ "step": 41
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018826815642458103,
+ "loss": 2.1731,
+ "step": 42
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.00018798882681564247,
+ "loss": 2.2373,
+ "step": 43
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0001877094972067039,
+ "loss": 2.2295,
+ "step": 44
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.00018743016759776538,
+ "loss": 2.1947,
+ "step": 45
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018715083798882682,
+ "loss": 2.2115,
+ "step": 46
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018687150837988828,
+ "loss": 2.1224,
+ "step": 47
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018659217877094972,
+ "loss": 2.2137,
+ "step": 48
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018631284916201116,
+ "loss": 2.2338,
+ "step": 49
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.00018603351955307266,
+ "loss": 2.1298,
+ "step": 50
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0001857541899441341,
+ "loss": 2.0883,
+ "step": 51
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.00018547486033519553,
+ "loss": 2.1216,
+ "step": 52
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.000185195530726257,
+ "loss": 2.2112,
+ "step": 53
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.00018491620111731844,
+ "loss": 2.1224,
+ "step": 54
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0001846368715083799,
+ "loss": 2.2375,
+ "step": 55
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.00018435754189944135,
+ "loss": 2.2235,
+ "step": 56
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018407821229050279,
+ "loss": 2.1682,
+ "step": 57
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018379888268156425,
+ "loss": 2.2077,
+ "step": 58
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.00018351955307262572,
+ "loss": 2.1596,
+ "step": 59
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018324022346368716,
+ "loss": 2.1311,
+ "step": 60
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018296089385474862,
+ "loss": 2.1333,
+ "step": 61
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018268156424581006,
+ "loss": 2.0901,
+ "step": 62
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018240223463687153,
+ "loss": 2.1971,
+ "step": 63
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.00018212290502793297,
+ "loss": 2.2602,
+ "step": 64
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0001818435754189944,
+ "loss": 2.2194,
+ "step": 65
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018156424581005588,
+ "loss": 2.1218,
+ "step": 66
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018128491620111731,
+ "loss": 2.2049,
+ "step": 67
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.00018100558659217878,
+ "loss": 2.1521,
+ "step": 68
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.00018072625698324025,
+ "loss": 2.112,
+ "step": 69
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0001804469273743017,
+ "loss": 2.1906,
+ "step": 70
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.00018016759776536313,
+ "loss": 2.1717,
+ "step": 71
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0001798882681564246,
+ "loss": 2.0712,
+ "step": 72
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.00017960893854748603,
+ "loss": 2.141,
+ "step": 73
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0001793296089385475,
+ "loss": 2.0656,
+ "step": 74
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017905027932960894,
+ "loss": 2.1125,
+ "step": 75
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017877094972067038,
+ "loss": 2.0869,
+ "step": 76
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.00017849162011173187,
+ "loss": 2.2478,
+ "step": 77
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0001782122905027933,
+ "loss": 2.1535,
+ "step": 78
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.00017793296089385475,
+ "loss": 2.1927,
+ "step": 79
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017765363128491622,
+ "loss": 2.1213,
+ "step": 80
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017737430167597766,
+ "loss": 2.0981,
+ "step": 81
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017709497206703912,
+ "loss": 2.1828,
+ "step": 82
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017681564245810056,
+ "loss": 2.0562,
+ "step": 83
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.000176536312849162,
+ "loss": 2.1334,
+ "step": 84
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.00017625698324022347,
+ "loss": 2.1225,
+ "step": 85
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.00017597765363128493,
+ "loss": 2.2098,
+ "step": 86
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017569832402234637,
+ "loss": 2.1519,
+ "step": 87
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017541899441340784,
+ "loss": 2.1132,
+ "step": 88
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017513966480446928,
+ "loss": 2.0333,
+ "step": 89
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017486033519553075,
+ "loss": 2.2764,
+ "step": 90
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017458100558659218,
+ "loss": 2.1838,
+ "step": 91
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017430167597765362,
+ "loss": 2.1386,
+ "step": 92
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0001740223463687151,
+ "loss": 2.1034,
+ "step": 93
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.00017374301675977656,
+ "loss": 2.0346,
+ "step": 94
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.000173463687150838,
+ "loss": 2.0274,
+ "step": 95
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.00017318435754189946,
+ "loss": 2.1036,
+ "step": 96
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0001729050279329609,
+ "loss": 2.1208,
+ "step": 97
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.00017262569832402237,
+ "loss": 2.0572,
+ "step": 98
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0001723463687150838,
+ "loss": 2.1702,
+ "step": 99
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.00017206703910614525,
+ "loss": 2.1302,
+ "step": 100
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0001717877094972067,
+ "loss": 2.0175,
+ "step": 101
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.00017150837988826815,
+ "loss": 2.1006,
+ "step": 102
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017122905027932962,
+ "loss": 2.0662,
+ "step": 103
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017094972067039109,
+ "loss": 1.988,
+ "step": 104
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017067039106145253,
+ "loss": 2.1008,
+ "step": 105
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017039106145251396,
+ "loss": 2.1482,
+ "step": 106
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00017011173184357543,
+ "loss": 2.1052,
+ "step": 107
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00016983240223463687,
+ "loss": 2.0978,
+ "step": 108
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016955307262569834,
+ "loss": 2.1303,
+ "step": 109
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016927374301675978,
+ "loss": 2.0794,
+ "step": 110
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.00016899441340782122,
+ "loss": 2.1059,
+ "step": 111
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0001687150837988827,
+ "loss": 1.9642,
+ "step": 112
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.00016843575418994415,
+ "loss": 2.0415,
+ "step": 113
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0001681564245810056,
+ "loss": 2.0795,
+ "step": 114
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.00016787709497206705,
+ "loss": 2.0238,
+ "step": 115
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0001675977653631285,
+ "loss": 2.078,
+ "step": 116
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.00016731843575418996,
+ "loss": 2.1362,
+ "step": 117
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0001670391061452514,
+ "loss": 2.0552,
+ "step": 118
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.00016675977653631284,
+ "loss": 2.213,
+ "step": 119
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0001664804469273743,
+ "loss": 1.9999,
+ "step": 120
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.00016620111731843577,
+ "loss": 2.06,
+ "step": 121
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0001659217877094972,
+ "loss": 2.0177,
+ "step": 122
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016564245810055868,
+ "loss": 2.0504,
+ "step": 123
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016536312849162012,
+ "loss": 2.0585,
+ "step": 124
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016508379888268158,
+ "loss": 2.0273,
+ "step": 125
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016480446927374302,
+ "loss": 2.0549,
+ "step": 126
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.00016452513966480446,
+ "loss": 2.044,
+ "step": 127
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016424581005586593,
+ "loss": 2.0731,
+ "step": 128
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016396648044692737,
+ "loss": 2.0568,
+ "step": 129
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.00016368715083798883,
+ "loss": 2.007,
+ "step": 130
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0001634078212290503,
+ "loss": 2.0707,
+ "step": 131
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.00016312849162011174,
+ "loss": 1.9793,
+ "step": 132
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0001628491620111732,
+ "loss": 2.1311,
+ "step": 133
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016256983240223465,
+ "loss": 2.0016,
+ "step": 134
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016229050279329609,
+ "loss": 1.9945,
+ "step": 135
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.00016201117318435755,
+ "loss": 2.0186,
+ "step": 136
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.000161731843575419,
+ "loss": 2.0971,
+ "step": 137
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.00016145251396648046,
+ "loss": 2.0883,
+ "step": 138
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016117318435754192,
+ "loss": 2.0803,
+ "step": 139
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016089385474860336,
+ "loss": 2.0617,
+ "step": 140
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016061452513966483,
+ "loss": 2.1265,
+ "step": 141
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016033519553072627,
+ "loss": 2.0151,
+ "step": 142
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0001600558659217877,
+ "loss": 1.996,
+ "step": 143
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.00015977653631284918,
+ "loss": 2.0164,
+ "step": 144
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.00015949720670391061,
+ "loss": 2.0314,
+ "step": 145
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015921787709497208,
+ "loss": 1.9501,
+ "step": 146
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015893854748603352,
+ "loss": 2.087,
+ "step": 147
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.000158659217877095,
+ "loss": 2.0262,
+ "step": 148
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.00015837988826815643,
+ "loss": 2.0765,
+ "step": 149
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0001581005586592179,
+ "loss": 2.105,
+ "step": 150
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.00015782122905027933,
+ "loss": 1.9863,
+ "step": 151
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0001575418994413408,
+ "loss": 1.9873,
+ "step": 152
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.00015726256983240224,
+ "loss": 2.0094,
+ "step": 153
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.00015698324022346368,
+ "loss": 1.9141,
+ "step": 154
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015670391061452514,
+ "loss": 1.917,
+ "step": 155
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015642458100558658,
+ "loss": 2.109,
+ "step": 156
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015614525139664805,
+ "loss": 1.9799,
+ "step": 157
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015586592178770952,
+ "loss": 1.9571,
+ "step": 158
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015558659217877096,
+ "loss": 1.9931,
+ "step": 159
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015530726256983242,
+ "loss": 2.1004,
+ "step": 160
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.00015502793296089386,
+ "loss": 2.0385,
+ "step": 161
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0001547486033519553,
+ "loss": 1.9751,
+ "step": 162
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.00015446927374301677,
+ "loss": 2.0544,
+ "step": 163
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0001541899441340782,
+ "loss": 2.0069,
+ "step": 164
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.00015391061452513967,
+ "loss": 1.9576,
+ "step": 165
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015363128491620114,
+ "loss": 1.8991,
+ "step": 166
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015335195530726258,
+ "loss": 1.9336,
+ "step": 167
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015307262569832405,
+ "loss": 1.9736,
+ "step": 168
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015279329608938548,
+ "loss": 1.9702,
+ "step": 169
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.00015251396648044692,
+ "loss": 1.9055,
+ "step": 170
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0001522346368715084,
+ "loss": 2.0503,
+ "step": 171
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.00015195530726256983,
+ "loss": 2.0039,
+ "step": 172
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0001516759776536313,
+ "loss": 1.9406,
+ "step": 173
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.00015139664804469274,
+ "loss": 2.0525,
+ "step": 174
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0001511173184357542,
+ "loss": 1.9234,
+ "step": 175
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.00015083798882681567,
+ "loss": 1.8614,
+ "step": 176
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0001505586592178771,
+ "loss": 1.9616,
+ "step": 177
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.00015027932960893855,
+ "loss": 1.9509,
+ "step": 178
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.9592,
+ "step": 179
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014972067039106145,
+ "loss": 1.8991,
+ "step": 180
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014944134078212292,
+ "loss": 1.9127,
+ "step": 181
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.00014916201117318436,
+ "loss": 1.8982,
+ "step": 182
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0001488826815642458,
+ "loss": 1.9534,
+ "step": 183
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0001486033519553073,
+ "loss": 1.7794,
+ "step": 184
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.00014832402234636873,
+ "loss": 1.7958,
+ "step": 185
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014804469273743017,
+ "loss": 1.8282,
+ "step": 186
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014776536312849164,
+ "loss": 2.0423,
+ "step": 187
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.00014748603351955308,
+ "loss": 1.9282,
+ "step": 188
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014720670391061454,
+ "loss": 1.9072,
+ "step": 189
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014692737430167598,
+ "loss": 1.8665,
+ "step": 190
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.00014664804469273742,
+ "loss": 1.9021,
+ "step": 191
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.0001463687150837989,
+ "loss": 1.7308,
+ "step": 192
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.00014608938547486035,
+ "loss": 1.9165,
+ "step": 193
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.0001458100558659218,
+ "loss": 1.842,
+ "step": 194
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.00014553072625698326,
+ "loss": 1.9128,
+ "step": 195
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.0001452513966480447,
+ "loss": 1.8005,
+ "step": 196
+ },
+ {
+ "epoch": 1.1,
+ "learning_rate": 0.00014497206703910614,
+ "loss": 1.8547,
+ "step": 197
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.0001446927374301676,
+ "loss": 1.9042,
+ "step": 198
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.00014441340782122905,
+ "loss": 1.8609,
+ "step": 199
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.0001441340782122905,
+ "loss": 1.9591,
+ "step": 200
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.00014385474860335195,
+ "loss": 1.8722,
+ "step": 201
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014357541899441342,
+ "loss": 1.8535,
+ "step": 202
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014329608938547488,
+ "loss": 1.8676,
+ "step": 203
+ },
+ {
+ "epoch": 1.14,
+ "learning_rate": 0.00014301675977653632,
+ "loss": 1.8976,
+ "step": 204
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014273743016759776,
+ "loss": 1.7723,
+ "step": 205
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014245810055865923,
+ "loss": 1.8165,
+ "step": 206
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014217877094972067,
+ "loss": 1.7811,
+ "step": 207
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014189944134078214,
+ "loss": 1.908,
+ "step": 208
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014162011173184357,
+ "loss": 1.7663,
+ "step": 209
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014134078212290501,
+ "loss": 1.6779,
+ "step": 210
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.0001410614525139665,
+ "loss": 1.9039,
+ "step": 211
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.00014078212290502795,
+ "loss": 1.8033,
+ "step": 212
+ },
+ {
+ "epoch": 1.19,
+ "learning_rate": 0.00014050279329608939,
+ "loss": 1.8251,
+ "step": 213
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.00014022346368715085,
+ "loss": 1.8505,
+ "step": 214
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.0001399441340782123,
+ "loss": 1.8147,
+ "step": 215
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.00013966480446927376,
+ "loss": 1.8419,
+ "step": 216
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.0001393854748603352,
+ "loss": 1.8401,
+ "step": 217
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.00013910614525139664,
+ "loss": 1.8912,
+ "step": 218
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.0001388268156424581,
+ "loss": 1.7548,
+ "step": 219
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.00013854748603351957,
+ "loss": 1.8741,
+ "step": 220
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.000138268156424581,
+ "loss": 1.9549,
+ "step": 221
+ },
+ {
+ "epoch": 1.24,
+ "learning_rate": 0.00013798882681564248,
+ "loss": 1.9093,
+ "step": 222
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013770949720670392,
+ "loss": 1.7896,
+ "step": 223
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013743016759776538,
+ "loss": 1.8491,
+ "step": 224
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013715083798882682,
+ "loss": 1.7851,
+ "step": 225
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013687150837988826,
+ "loss": 1.6992,
+ "step": 226
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013659217877094973,
+ "loss": 1.9765,
+ "step": 227
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013631284916201117,
+ "loss": 1.8179,
+ "step": 228
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.00013603351955307263,
+ "loss": 1.8548,
+ "step": 229
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.0001357541899441341,
+ "loss": 1.8843,
+ "step": 230
+ },
+ {
+ "epoch": 1.29,
+ "learning_rate": 0.00013547486033519554,
+ "loss": 1.9105,
+ "step": 231
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013519553072625698,
+ "loss": 1.8748,
+ "step": 232
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013491620111731844,
+ "loss": 1.7976,
+ "step": 233
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013463687150837988,
+ "loss": 1.7369,
+ "step": 234
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013435754189944135,
+ "loss": 1.7808,
+ "step": 235
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.0001340782122905028,
+ "loss": 1.8385,
+ "step": 236
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.00013379888268156423,
+ "loss": 1.8295,
+ "step": 237
+ },
+ {
+ "epoch": 1.33,
+ "learning_rate": 0.00013351955307262572,
+ "loss": 1.757,
+ "step": 238
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.00013324022346368716,
+ "loss": 1.7904,
+ "step": 239
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.0001329608938547486,
+ "loss": 1.7632,
+ "step": 240
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.00013268156424581007,
+ "loss": 1.7867,
+ "step": 241
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.0001324022346368715,
+ "loss": 1.8259,
+ "step": 242
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.00013212290502793297,
+ "loss": 1.6655,
+ "step": 243
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.0001318435754189944,
+ "loss": 1.6848,
+ "step": 244
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013156424581005585,
+ "loss": 1.7931,
+ "step": 245
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013128491620111732,
+ "loss": 1.7868,
+ "step": 246
+ },
+ {
+ "epoch": 1.38,
+ "learning_rate": 0.00013100558659217879,
+ "loss": 1.7732,
+ "step": 247
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.00013072625698324022,
+ "loss": 1.7851,
+ "step": 248
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.0001304469273743017,
+ "loss": 1.7406,
+ "step": 249
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.00013016759776536313,
+ "loss": 1.5853,
+ "step": 250
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.0001298882681564246,
+ "loss": 1.8271,
+ "step": 251
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012960893854748604,
+ "loss": 1.6054,
+ "step": 252
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012932960893854748,
+ "loss": 1.6884,
+ "step": 253
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012905027932960894,
+ "loss": 1.7333,
+ "step": 254
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012877094972067038,
+ "loss": 1.803,
+ "step": 255
+ },
+ {
+ "epoch": 1.43,
+ "learning_rate": 0.00012849162011173185,
+ "loss": 1.63,
+ "step": 256
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012821229050279331,
+ "loss": 1.8023,
+ "step": 257
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012793296089385475,
+ "loss": 1.762,
+ "step": 258
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012765363128491622,
+ "loss": 1.7565,
+ "step": 259
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012737430167597766,
+ "loss": 1.5937,
+ "step": 260
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.0001270949720670391,
+ "loss": 1.7208,
+ "step": 261
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.00012681564245810057,
+ "loss": 1.8097,
+ "step": 262
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.000126536312849162,
+ "loss": 1.7042,
+ "step": 263
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.00012625698324022347,
+ "loss": 1.7892,
+ "step": 264
+ },
+ {
+ "epoch": 1.48,
+ "learning_rate": 0.00012597765363128494,
+ "loss": 1.752,
+ "step": 265
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012569832402234638,
+ "loss": 1.7333,
+ "step": 266
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012541899441340784,
+ "loss": 1.7492,
+ "step": 267
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012513966480446928,
+ "loss": 1.818,
+ "step": 268
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012486033519553072,
+ "loss": 1.7146,
+ "step": 269
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.0001245810055865922,
+ "loss": 1.7958,
+ "step": 270
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.00012430167597765363,
+ "loss": 1.7212,
+ "step": 271
+ },
+ {
+ "epoch": 1.52,
+ "learning_rate": 0.0001240223463687151,
+ "loss": 1.8835,
+ "step": 272
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.00012374301675977656,
+ "loss": 1.7486,
+ "step": 273
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.000123463687150838,
+ "loss": 1.8103,
+ "step": 274
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.00012318435754189944,
+ "loss": 1.7557,
+ "step": 275
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.0001229050279329609,
+ "loss": 1.6198,
+ "step": 276
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.00012262569832402235,
+ "loss": 1.6971,
+ "step": 277
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.0001223463687150838,
+ "loss": 1.668,
+ "step": 278
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.00012206703910614525,
+ "loss": 1.8795,
+ "step": 279
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.0001217877094972067,
+ "loss": 1.6412,
+ "step": 280
+ },
+ {
+ "epoch": 1.57,
+ "learning_rate": 0.00012150837988826816,
+ "loss": 1.7497,
+ "step": 281
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012122905027932962,
+ "loss": 1.5577,
+ "step": 282
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012094972067039108,
+ "loss": 1.8049,
+ "step": 283
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012067039106145253,
+ "loss": 1.6834,
+ "step": 284
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012039106145251397,
+ "loss": 1.7978,
+ "step": 285
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00012011173184357542,
+ "loss": 1.6558,
+ "step": 286
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00011983240223463687,
+ "loss": 1.7561,
+ "step": 287
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011955307262569833,
+ "loss": 1.6807,
+ "step": 288
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011927374301675978,
+ "loss": 1.6937,
+ "step": 289
+ },
+ {
+ "epoch": 1.62,
+ "learning_rate": 0.00011899441340782122,
+ "loss": 1.6591,
+ "step": 290
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.0001187150837988827,
+ "loss": 1.6771,
+ "step": 291
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.00011843575418994415,
+ "loss": 1.7743,
+ "step": 292
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011815642458100559,
+ "loss": 1.5857,
+ "step": 293
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011787709497206705,
+ "loss": 1.6999,
+ "step": 294
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.0001175977653631285,
+ "loss": 1.5661,
+ "step": 295
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.00011731843575418995,
+ "loss": 1.7235,
+ "step": 296
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.0001170391061452514,
+ "loss": 1.607,
+ "step": 297
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.00011675977653631284,
+ "loss": 1.68,
+ "step": 298
+ },
+ {
+ "epoch": 1.67,
+ "learning_rate": 0.0001164804469273743,
+ "loss": 1.6938,
+ "step": 299
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011620111731843578,
+ "loss": 1.6315,
+ "step": 300
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011592178770949722,
+ "loss": 1.6802,
+ "step": 301
+ },
+ {
+ "epoch": 1.69,
+ "learning_rate": 0.00011564245810055867,
+ "loss": 1.7174,
+ "step": 302
+ },
+ {
+ "epoch": 1.69,
+ "learning_rate": 0.00011536312849162012,
+ "loss": 1.5212,
+ "step": 303
+ },
+ {
+ "epoch": 1.7,
+ "learning_rate": 0.00011508379888268157,
+ "loss": 1.5808,
+ "step": 304
+ },
+ {
+ "epoch": 1.7,
+ "learning_rate": 0.00011480446927374303,
+ "loss": 1.6152,
+ "step": 305
+ },
+ {
+ "epoch": 1.71,
+ "learning_rate": 0.00011452513966480447,
+ "loss": 1.5435,
+ "step": 306
+ },
+ {
+ "epoch": 1.72,
+ "learning_rate": 0.00011424581005586592,
+ "loss": 1.6603,
+ "step": 307
+ },
+ {
+ "epoch": 1.72,
+ "learning_rate": 0.00011396648044692737,
+ "loss": 1.685,
+ "step": 308
+ },
+ {
+ "epoch": 1.73,
+ "learning_rate": 0.00011368715083798884,
+ "loss": 1.6002,
+ "step": 309
+ },
+ {
+ "epoch": 1.73,
+ "learning_rate": 0.00011340782122905029,
+ "loss": 1.6046,
+ "step": 310
+ },
+ {
+ "epoch": 1.74,
+ "learning_rate": 0.00011312849162011174,
+ "loss": 1.5969,
+ "step": 311
+ },
+ {
+ "epoch": 1.74,
+ "learning_rate": 0.0001128491620111732,
+ "loss": 1.5845,
+ "step": 312
+ },
+ {
+ "epoch": 1.75,
+ "learning_rate": 0.00011256983240223464,
+ "loss": 1.8183,
+ "step": 313
+ },
+ {
+ "epoch": 1.75,
+ "learning_rate": 0.00011229050279329609,
+ "loss": 1.6953,
+ "step": 314
+ },
+ {
+ "epoch": 1.76,
+ "learning_rate": 0.00011201117318435754,
+ "loss": 1.7787,
+ "step": 315
+ },
+ {
+ "epoch": 1.77,
+ "learning_rate": 0.000111731843575419,
+ "loss": 1.6422,
+ "step": 316
+ },
+ {
+ "epoch": 1.77,
+ "learning_rate": 0.00011145251396648045,
+ "loss": 1.7034,
+ "step": 317
+ },
+ {
+ "epoch": 1.78,
+ "learning_rate": 0.00011117318435754192,
+ "loss": 1.7301,
+ "step": 318
+ },
+ {
+ "epoch": 1.78,
+ "learning_rate": 0.00011089385474860337,
+ "loss": 1.7084,
+ "step": 319
+ },
+ {
+ "epoch": 1.79,
+ "learning_rate": 0.00011061452513966482,
+ "loss": 1.772,
+ "step": 320
+ },
+ {
+ "epoch": 1.79,
+ "learning_rate": 0.00011033519553072626,
+ "loss": 1.5733,
+ "step": 321
+ },
+ {
+ "epoch": 1.8,
+ "learning_rate": 0.00011005586592178771,
+ "loss": 1.6423,
+ "step": 322
+ },
+ {
+ "epoch": 1.8,
+ "learning_rate": 0.00010977653631284917,
+ "loss": 1.5809,
+ "step": 323
+ },
+ {
+ "epoch": 1.81,
+ "learning_rate": 0.00010949720670391062,
+ "loss": 1.6781,
+ "step": 324
+ },
+ {
+ "epoch": 1.82,
+ "learning_rate": 0.00010921787709497207,
+ "loss": 1.6788,
+ "step": 325
+ },
+ {
+ "epoch": 1.82,
+ "learning_rate": 0.00010893854748603351,
+ "loss": 1.6346,
+ "step": 326
+ },
+ {
+ "epoch": 1.83,
+ "learning_rate": 0.00010865921787709499,
+ "loss": 1.6634,
+ "step": 327
+ },
+ {
+ "epoch": 1.83,
+ "learning_rate": 0.00010837988826815643,
+ "loss": 1.7561,
+ "step": 328
+ },
+ {
+ "epoch": 1.84,
+ "learning_rate": 0.00010810055865921788,
+ "loss": 1.66,
+ "step": 329
+ },
+ {
+ "epoch": 1.84,
+ "learning_rate": 0.00010782122905027934,
+ "loss": 1.7298,
+ "step": 330
+ },
+ {
+ "epoch": 1.85,
+ "learning_rate": 0.00010754189944134079,
+ "loss": 1.6893,
+ "step": 331
+ },
+ {
+ "epoch": 1.85,
+ "learning_rate": 0.00010726256983240224,
+ "loss": 1.7631,
+ "step": 332
+ },
+ {
+ "epoch": 1.86,
+ "learning_rate": 0.00010698324022346368,
+ "loss": 1.6633,
+ "step": 333
+ },
+ {
+ "epoch": 1.87,
+ "learning_rate": 0.00010670391061452513,
+ "loss": 1.5388,
+ "step": 334
+ },
+ {
+ "epoch": 1.87,
+ "learning_rate": 0.00010642458100558659,
+ "loss": 1.6718,
+ "step": 335
+ },
+ {
+ "epoch": 1.88,
+ "learning_rate": 0.00010614525139664805,
+ "loss": 1.5536,
+ "step": 336
+ },
+ {
+ "epoch": 1.88,
+ "learning_rate": 0.00010586592178770951,
+ "loss": 1.6483,
+ "step": 337
+ },
+ {
+ "epoch": 1.89,
+ "learning_rate": 0.00010558659217877096,
+ "loss": 1.5774,
+ "step": 338
+ },
+ {
+ "epoch": 1.89,
+ "learning_rate": 0.00010530726256983241,
+ "loss": 1.6366,
+ "step": 339
+ },
+ {
+ "epoch": 1.9,
+ "learning_rate": 0.00010502793296089387,
+ "loss": 1.5567,
+ "step": 340
+ },
+ {
+ "epoch": 1.91,
+ "learning_rate": 0.0001047486033519553,
+ "loss": 1.5323,
+ "step": 341
+ },
+ {
+ "epoch": 1.91,
+ "learning_rate": 0.00010446927374301676,
+ "loss": 1.4608,
+ "step": 342
+ },
+ {
+ "epoch": 1.92,
+ "learning_rate": 0.00010418994413407821,
+ "loss": 1.5933,
+ "step": 343
+ },
+ {
+ "epoch": 1.92,
+ "learning_rate": 0.00010391061452513966,
+ "loss": 1.6625,
+ "step": 344
+ },
+ {
+ "epoch": 1.93,
+ "learning_rate": 0.00010363128491620113,
+ "loss": 1.7236,
+ "step": 345
+ },
+ {
+ "epoch": 1.93,
+ "learning_rate": 0.00010335195530726258,
+ "loss": 1.759,
+ "step": 346
+ },
+ {
+ "epoch": 1.94,
+ "learning_rate": 0.00010307262569832404,
+ "loss": 1.7248,
+ "step": 347
+ },
+ {
+ "epoch": 1.94,
+ "learning_rate": 0.00010279329608938548,
+ "loss": 1.5144,
+ "step": 348
+ },
+ {
+ "epoch": 1.95,
+ "learning_rate": 0.00010251396648044693,
+ "loss": 1.6905,
+ "step": 349
+ },
+ {
+ "epoch": 1.96,
+ "learning_rate": 0.00010223463687150838,
+ "loss": 1.6119,
+ "step": 350
+ },
+ {
+ "epoch": 1.96,
+ "learning_rate": 0.00010195530726256983,
+ "loss": 1.5464,
+ "step": 351
+ },
+ {
+ "epoch": 1.97,
+ "learning_rate": 0.00010167597765363129,
+ "loss": 1.6901,
+ "step": 352
+ },
+ {
+ "epoch": 1.97,
+ "learning_rate": 0.00010139664804469273,
+ "loss": 1.3511,
+ "step": 353
+ },
+ {
+ "epoch": 1.98,
+ "learning_rate": 0.0001011173184357542,
+ "loss": 1.5434,
+ "step": 354
+ },
+ {
+ "epoch": 1.98,
+ "learning_rate": 0.00010083798882681566,
+ "loss": 1.5891,
+ "step": 355
+ },
+ {
+ "epoch": 1.99,
+ "learning_rate": 0.0001005586592178771,
+ "loss": 1.6658,
+ "step": 356
+ },
+ {
+ "epoch": 1.99,
+ "learning_rate": 0.00010027932960893855,
+ "loss": 1.5657,
+ "step": 357
+ },
+ {
+ "epoch": 2.0,
+ "learning_rate": 0.0001,
+ "loss": 1.7005,
+ "step": 358
+ },
+ {
+ "epoch": 2.01,
+ "learning_rate": 9.972067039106146e-05,
+ "loss": 1.4202,
+ "step": 359
+ },
+ {
+ "epoch": 2.01,
+ "learning_rate": 9.944134078212291e-05,
+ "loss": 1.5262,
+ "step": 360
+ },
+ {
+ "epoch": 2.02,
+ "learning_rate": 9.916201117318436e-05,
+ "loss": 1.6323,
+ "step": 361
+ },
+ {
+ "epoch": 2.02,
+ "learning_rate": 9.888268156424582e-05,
+ "loss": 1.5521,
+ "step": 362
+ },
+ {
+ "epoch": 2.03,
+ "learning_rate": 9.860335195530727e-05,
+ "loss": 1.5762,
+ "step": 363
+ },
+ {
+ "epoch": 2.03,
+ "learning_rate": 9.832402234636872e-05,
+ "loss": 1.613,
+ "step": 364
+ },
+ {
+ "epoch": 2.04,
+ "learning_rate": 9.804469273743018e-05,
+ "loss": 1.4231,
+ "step": 365
+ },
+ {
+ "epoch": 2.04,
+ "learning_rate": 9.776536312849163e-05,
+ "loss": 1.5706,
+ "step": 366
+ },
+ {
+ "epoch": 2.05,
+ "learning_rate": 9.748603351955308e-05,
+ "loss": 1.5245,
+ "step": 367
+ },
+ {
+ "epoch": 2.06,
+ "learning_rate": 9.720670391061453e-05,
+ "loss": 1.4771,
+ "step": 368
+ },
+ {
+ "epoch": 2.06,
+ "learning_rate": 9.692737430167597e-05,
+ "loss": 1.596,
+ "step": 369
+ },
+ {
+ "epoch": 2.07,
+ "learning_rate": 9.664804469273744e-05,
+ "loss": 1.537,
+ "step": 370
+ },
+ {
+ "epoch": 2.07,
+ "learning_rate": 9.636871508379889e-05,
+ "loss": 1.4276,
+ "step": 371
+ },
+ {
+ "epoch": 2.08,
+ "learning_rate": 9.608938547486033e-05,
+ "loss": 1.4746,
+ "step": 372
+ },
+ {
+ "epoch": 2.08,
+ "learning_rate": 9.581005586592178e-05,
+ "loss": 1.4374,
+ "step": 373
+ },
+ {
+ "epoch": 2.09,
+ "learning_rate": 9.553072625698325e-05,
+ "loss": 1.4704,
+ "step": 374
+ },
+ {
+ "epoch": 2.09,
+ "learning_rate": 9.52513966480447e-05,
+ "loss": 1.5997,
+ "step": 375
+ },
+ {
+ "epoch": 2.1,
+ "learning_rate": 9.497206703910614e-05,
+ "loss": 1.5034,
+ "step": 376
+ },
+ {
+ "epoch": 2.11,
+ "learning_rate": 9.46927374301676e-05,
+ "loss": 1.6392,
+ "step": 377
+ },
+ {
+ "epoch": 2.11,
+ "learning_rate": 9.441340782122905e-05,
+ "loss": 1.5611,
+ "step": 378
+ },
+ {
+ "epoch": 2.12,
+ "learning_rate": 9.413407821229052e-05,
+ "loss": 1.3384,
+ "step": 379
+ },
+ {
+ "epoch": 2.12,
+ "learning_rate": 9.385474860335196e-05,
+ "loss": 1.5732,
+ "step": 380
+ },
+ {
+ "epoch": 2.13,
+ "learning_rate": 9.357541899441341e-05,
+ "loss": 1.3874,
+ "step": 381
+ },
+ {
+ "epoch": 2.13,
+ "learning_rate": 9.329608938547486e-05,
+ "loss": 1.4139,
+ "step": 382
+ },
+ {
+ "epoch": 2.14,
+ "learning_rate": 9.301675977653633e-05,
+ "loss": 1.5195,
+ "step": 383
+ },
+ {
+ "epoch": 2.15,
+ "learning_rate": 9.273743016759777e-05,
+ "loss": 1.4371,
+ "step": 384
+ },
+ {
+ "epoch": 2.15,
+ "learning_rate": 9.245810055865922e-05,
+ "loss": 1.4411,
+ "step": 385
+ },
+ {
+ "epoch": 2.16,
+ "learning_rate": 9.217877094972067e-05,
+ "loss": 1.5214,
+ "step": 386
+ },
+ {
+ "epoch": 2.16,
+ "learning_rate": 9.189944134078213e-05,
+ "loss": 1.4971,
+ "step": 387
+ },
+ {
+ "epoch": 2.17,
+ "learning_rate": 9.162011173184358e-05,
+ "loss": 1.3453,
+ "step": 388
+ },
+ {
+ "epoch": 2.17,
+ "learning_rate": 9.134078212290503e-05,
+ "loss": 1.394,
+ "step": 389
+ },
+ {
+ "epoch": 2.18,
+ "learning_rate": 9.106145251396648e-05,
+ "loss": 1.5058,
+ "step": 390
+ },
+ {
+ "epoch": 2.18,
+ "learning_rate": 9.078212290502794e-05,
+ "loss": 1.4855,
+ "step": 391
+ },
+ {
+ "epoch": 2.19,
+ "learning_rate": 9.050279329608939e-05,
+ "loss": 1.4647,
+ "step": 392
+ },
+ {
+ "epoch": 2.2,
+ "learning_rate": 9.022346368715084e-05,
+ "loss": 1.3435,
+ "step": 393
+ },
+ {
+ "epoch": 2.2,
+ "learning_rate": 8.99441340782123e-05,
+ "loss": 1.5815,
+ "step": 394
+ },
+ {
+ "epoch": 2.21,
+ "learning_rate": 8.966480446927375e-05,
+ "loss": 1.4742,
+ "step": 395
+ },
+ {
+ "epoch": 2.21,
+ "learning_rate": 8.938547486033519e-05,
+ "loss": 1.5389,
+ "step": 396
+ },
+ {
+ "epoch": 2.22,
+ "learning_rate": 8.910614525139666e-05,
+ "loss": 1.384,
+ "step": 397
+ },
+ {
+ "epoch": 2.22,
+ "learning_rate": 8.882681564245811e-05,
+ "loss": 1.3967,
+ "step": 398
+ },
+ {
+ "epoch": 2.23,
+ "learning_rate": 8.854748603351956e-05,
+ "loss": 1.442,
+ "step": 399
+ },
+ {
+ "epoch": 2.23,
+ "learning_rate": 8.8268156424581e-05,
+ "loss": 1.396,
+ "step": 400
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 716,
+ "num_train_epochs": 4,
+ "save_steps": 100,
+ "total_flos": 2.056923590572032e+17,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-400/training_args.bin b/checkpoint-400/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8672c716e925d0028b4938db147703f58656ff7
--- /dev/null
+++ b/checkpoint-400/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:188ae1c421cc0c6435d1f71d8d3423ac4abc7dba0e6fc2efcbc4dbe77c741317
+size 4027
diff --git a/checkpoint-500/README.md b/checkpoint-500/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..2f257a448caef1c59022426ad3ec9ba80acb3821
--- /dev/null
+++ b/checkpoint-500/README.md
@@ -0,0 +1,60 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+### Framework versions
+
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+- PEFT 0.6.0.dev0
+
+- PEFT 0.6.0.dev0
diff --git a/checkpoint-500/adapter_config.json b/checkpoint-500/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c661f74e40394c079de68f07861d73c341803b67
--- /dev/null
+++ b/checkpoint-500/adapter_config.json
@@ -0,0 +1,25 @@
+{
+ "alpha_pattern": {},
+ "auto_mapping": null,
+ "base_model_name_or_path": "bigscience/bloomz-3b",
+ "bias": "none",
+ "fan_in_fan_out": false,
+ "inference_mode": true,
+ "init_lora_weights": true,
+ "layers_pattern": null,
+ "layers_to_transform": null,
+ "lora_alpha": 16,
+ "lora_dropout": 0.0,
+ "modules_to_save": null,
+ "peft_type": "LORA",
+ "r": 8,
+ "rank_pattern": {},
+ "revision": null,
+ "target_modules": [
+ "dense_4h_to_h",
+ "dense",
+ "dense_h_to_4h",
+ "query_key_value"
+ ],
+ "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/checkpoint-500/adapter_model.bin b/checkpoint-500/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..6198683070fc9df0fa08204fb385a9c9cf4fb8a2
--- /dev/null
+++ b/checkpoint-500/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbc5f8a93e505087cecc049ff476019df96de3cba21db5d67029ade06c844216
+size 39409357
diff --git a/checkpoint-500/optimizer.pt b/checkpoint-500/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..dec2baaffc2c1b79896eb5f10b68cc8630ce5592
--- /dev/null
+++ b/checkpoint-500/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbab141106d84467be5a7a978a93b8492dd00bcc4a497115fb4124bf8beaa046
+size 78844421
diff --git a/checkpoint-500/rng_state.pth b/checkpoint-500/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..e0df6592e27bef84a4beb9293e7b49666a4d652e
--- /dev/null
+++ b/checkpoint-500/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7508d4b8dd267de5cc58e972da25236687927651336a28f292c92f7f23951475
+size 14575
diff --git a/checkpoint-500/scheduler.pt b/checkpoint-500/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..53ba5c95b7d8f86bbe40d20004a07e2e74748d22
--- /dev/null
+++ b/checkpoint-500/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b841ff782e5772fca220707441f952d8fcec6aefcd65e69bf2fe08db08edfce1
+size 627
diff --git a/checkpoint-500/special_tokens_map.json b/checkpoint-500/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdafe480f024ff444c7492147536765ce5d55a2d
--- /dev/null
+++ b/checkpoint-500/special_tokens_map.json
@@ -0,0 +1,6 @@
+{
+ "bos_token": "",
+ "eos_token": "",
+ "pad_token": "",
+ "unk_token": ""
+}
diff --git a/checkpoint-500/tokenizer.json b/checkpoint-500/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..673c31abdeadf6576c3c754df86459e1ad64e207
--- /dev/null
+++ b/checkpoint-500/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85b00d7db4df5df2e3f01cacc3feda246002a672f3356eec7f4b04a22eb0dfbe
+size 14500570
diff --git a/checkpoint-500/tokenizer_config.json b/checkpoint-500/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b56cc9c2965c07132c35df3e2972e93d98c82c3
--- /dev/null
+++ b/checkpoint-500/tokenizer_config.json
@@ -0,0 +1,10 @@
+{
+ "add_prefix_space": false,
+ "bos_token": "",
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "",
+ "model_max_length": 1000000000000000019884624838656,
+ "pad_token": "",
+ "tokenizer_class": "BloomTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoint-500/trainer_state.json b/checkpoint-500/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..121cade410f9ee4e2a798857772481a6700d2a71
--- /dev/null
+++ b/checkpoint-500/trainer_state.json
@@ -0,0 +1,3019 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 2.793296089385475,
+ "eval_steps": 500,
+ "global_step": 500,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019972067039106145,
+ "loss": 2.6443,
+ "step": 1
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019944134078212292,
+ "loss": 2.4104,
+ "step": 2
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019916201117318435,
+ "loss": 2.4975,
+ "step": 3
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019888268156424582,
+ "loss": 2.3513,
+ "step": 4
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0001986033519553073,
+ "loss": 2.4274,
+ "step": 5
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.00019832402234636873,
+ "loss": 2.3628,
+ "step": 6
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0001980446927374302,
+ "loss": 2.3567,
+ "step": 7
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.00019776536312849163,
+ "loss": 2.4121,
+ "step": 8
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.00019748603351955307,
+ "loss": 2.4033,
+ "step": 9
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019720670391061454,
+ "loss": 2.2805,
+ "step": 10
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019692737430167598,
+ "loss": 2.2639,
+ "step": 11
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019664804469273744,
+ "loss": 2.2724,
+ "step": 12
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019636871508379888,
+ "loss": 2.332,
+ "step": 13
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019608938547486035,
+ "loss": 2.2261,
+ "step": 14
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019581005586592182,
+ "loss": 2.2208,
+ "step": 15
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.00019553072625698326,
+ "loss": 2.3351,
+ "step": 16
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0001952513966480447,
+ "loss": 2.2475,
+ "step": 17
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.00019497206703910616,
+ "loss": 2.3283,
+ "step": 18
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0001946927374301676,
+ "loss": 2.1346,
+ "step": 19
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.00019441340782122907,
+ "loss": 2.131,
+ "step": 20
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0001941340782122905,
+ "loss": 2.1718,
+ "step": 21
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.00019385474860335195,
+ "loss": 2.2446,
+ "step": 22
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0001935754189944134,
+ "loss": 2.306,
+ "step": 23
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.00019329608938547488,
+ "loss": 2.1908,
+ "step": 24
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.00019301675977653632,
+ "loss": 2.2844,
+ "step": 25
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019273743016759779,
+ "loss": 2.2235,
+ "step": 26
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019245810055865922,
+ "loss": 2.1842,
+ "step": 27
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019217877094972066,
+ "loss": 2.2675,
+ "step": 28
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019189944134078213,
+ "loss": 2.2532,
+ "step": 29
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019162011173184357,
+ "loss": 2.1788,
+ "step": 30
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019134078212290504,
+ "loss": 2.2494,
+ "step": 31
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0001910614525139665,
+ "loss": 2.1995,
+ "step": 32
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.00019078212290502794,
+ "loss": 2.1451,
+ "step": 33
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0001905027932960894,
+ "loss": 2.223,
+ "step": 34
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.00019022346368715085,
+ "loss": 2.2854,
+ "step": 35
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0001899441340782123,
+ "loss": 2.2265,
+ "step": 36
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.00018966480446927375,
+ "loss": 2.1214,
+ "step": 37
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0001893854748603352,
+ "loss": 2.1898,
+ "step": 38
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.00018910614525139666,
+ "loss": 2.1974,
+ "step": 39
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0001888268156424581,
+ "loss": 2.2259,
+ "step": 40
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018854748603351957,
+ "loss": 2.2094,
+ "step": 41
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018826815642458103,
+ "loss": 2.1731,
+ "step": 42
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.00018798882681564247,
+ "loss": 2.2373,
+ "step": 43
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0001877094972067039,
+ "loss": 2.2295,
+ "step": 44
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.00018743016759776538,
+ "loss": 2.1947,
+ "step": 45
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018715083798882682,
+ "loss": 2.2115,
+ "step": 46
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018687150837988828,
+ "loss": 2.1224,
+ "step": 47
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018659217877094972,
+ "loss": 2.2137,
+ "step": 48
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018631284916201116,
+ "loss": 2.2338,
+ "step": 49
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.00018603351955307266,
+ "loss": 2.1298,
+ "step": 50
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0001857541899441341,
+ "loss": 2.0883,
+ "step": 51
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.00018547486033519553,
+ "loss": 2.1216,
+ "step": 52
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.000185195530726257,
+ "loss": 2.2112,
+ "step": 53
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.00018491620111731844,
+ "loss": 2.1224,
+ "step": 54
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0001846368715083799,
+ "loss": 2.2375,
+ "step": 55
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.00018435754189944135,
+ "loss": 2.2235,
+ "step": 56
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018407821229050279,
+ "loss": 2.1682,
+ "step": 57
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018379888268156425,
+ "loss": 2.2077,
+ "step": 58
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.00018351955307262572,
+ "loss": 2.1596,
+ "step": 59
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018324022346368716,
+ "loss": 2.1311,
+ "step": 60
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018296089385474862,
+ "loss": 2.1333,
+ "step": 61
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018268156424581006,
+ "loss": 2.0901,
+ "step": 62
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018240223463687153,
+ "loss": 2.1971,
+ "step": 63
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.00018212290502793297,
+ "loss": 2.2602,
+ "step": 64
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0001818435754189944,
+ "loss": 2.2194,
+ "step": 65
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018156424581005588,
+ "loss": 2.1218,
+ "step": 66
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018128491620111731,
+ "loss": 2.2049,
+ "step": 67
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.00018100558659217878,
+ "loss": 2.1521,
+ "step": 68
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.00018072625698324025,
+ "loss": 2.112,
+ "step": 69
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0001804469273743017,
+ "loss": 2.1906,
+ "step": 70
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.00018016759776536313,
+ "loss": 2.1717,
+ "step": 71
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0001798882681564246,
+ "loss": 2.0712,
+ "step": 72
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.00017960893854748603,
+ "loss": 2.141,
+ "step": 73
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0001793296089385475,
+ "loss": 2.0656,
+ "step": 74
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017905027932960894,
+ "loss": 2.1125,
+ "step": 75
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017877094972067038,
+ "loss": 2.0869,
+ "step": 76
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.00017849162011173187,
+ "loss": 2.2478,
+ "step": 77
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0001782122905027933,
+ "loss": 2.1535,
+ "step": 78
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.00017793296089385475,
+ "loss": 2.1927,
+ "step": 79
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017765363128491622,
+ "loss": 2.1213,
+ "step": 80
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017737430167597766,
+ "loss": 2.0981,
+ "step": 81
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017709497206703912,
+ "loss": 2.1828,
+ "step": 82
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017681564245810056,
+ "loss": 2.0562,
+ "step": 83
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.000176536312849162,
+ "loss": 2.1334,
+ "step": 84
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.00017625698324022347,
+ "loss": 2.1225,
+ "step": 85
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.00017597765363128493,
+ "loss": 2.2098,
+ "step": 86
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017569832402234637,
+ "loss": 2.1519,
+ "step": 87
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017541899441340784,
+ "loss": 2.1132,
+ "step": 88
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017513966480446928,
+ "loss": 2.0333,
+ "step": 89
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017486033519553075,
+ "loss": 2.2764,
+ "step": 90
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017458100558659218,
+ "loss": 2.1838,
+ "step": 91
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017430167597765362,
+ "loss": 2.1386,
+ "step": 92
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0001740223463687151,
+ "loss": 2.1034,
+ "step": 93
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.00017374301675977656,
+ "loss": 2.0346,
+ "step": 94
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.000173463687150838,
+ "loss": 2.0274,
+ "step": 95
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.00017318435754189946,
+ "loss": 2.1036,
+ "step": 96
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0001729050279329609,
+ "loss": 2.1208,
+ "step": 97
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.00017262569832402237,
+ "loss": 2.0572,
+ "step": 98
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0001723463687150838,
+ "loss": 2.1702,
+ "step": 99
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.00017206703910614525,
+ "loss": 2.1302,
+ "step": 100
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0001717877094972067,
+ "loss": 2.0175,
+ "step": 101
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.00017150837988826815,
+ "loss": 2.1006,
+ "step": 102
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017122905027932962,
+ "loss": 2.0662,
+ "step": 103
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017094972067039109,
+ "loss": 1.988,
+ "step": 104
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017067039106145253,
+ "loss": 2.1008,
+ "step": 105
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017039106145251396,
+ "loss": 2.1482,
+ "step": 106
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00017011173184357543,
+ "loss": 2.1052,
+ "step": 107
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00016983240223463687,
+ "loss": 2.0978,
+ "step": 108
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016955307262569834,
+ "loss": 2.1303,
+ "step": 109
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016927374301675978,
+ "loss": 2.0794,
+ "step": 110
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.00016899441340782122,
+ "loss": 2.1059,
+ "step": 111
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0001687150837988827,
+ "loss": 1.9642,
+ "step": 112
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.00016843575418994415,
+ "loss": 2.0415,
+ "step": 113
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0001681564245810056,
+ "loss": 2.0795,
+ "step": 114
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.00016787709497206705,
+ "loss": 2.0238,
+ "step": 115
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0001675977653631285,
+ "loss": 2.078,
+ "step": 116
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.00016731843575418996,
+ "loss": 2.1362,
+ "step": 117
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0001670391061452514,
+ "loss": 2.0552,
+ "step": 118
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.00016675977653631284,
+ "loss": 2.213,
+ "step": 119
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0001664804469273743,
+ "loss": 1.9999,
+ "step": 120
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.00016620111731843577,
+ "loss": 2.06,
+ "step": 121
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0001659217877094972,
+ "loss": 2.0177,
+ "step": 122
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016564245810055868,
+ "loss": 2.0504,
+ "step": 123
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016536312849162012,
+ "loss": 2.0585,
+ "step": 124
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016508379888268158,
+ "loss": 2.0273,
+ "step": 125
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016480446927374302,
+ "loss": 2.0549,
+ "step": 126
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.00016452513966480446,
+ "loss": 2.044,
+ "step": 127
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016424581005586593,
+ "loss": 2.0731,
+ "step": 128
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016396648044692737,
+ "loss": 2.0568,
+ "step": 129
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.00016368715083798883,
+ "loss": 2.007,
+ "step": 130
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0001634078212290503,
+ "loss": 2.0707,
+ "step": 131
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.00016312849162011174,
+ "loss": 1.9793,
+ "step": 132
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0001628491620111732,
+ "loss": 2.1311,
+ "step": 133
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016256983240223465,
+ "loss": 2.0016,
+ "step": 134
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016229050279329609,
+ "loss": 1.9945,
+ "step": 135
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.00016201117318435755,
+ "loss": 2.0186,
+ "step": 136
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.000161731843575419,
+ "loss": 2.0971,
+ "step": 137
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.00016145251396648046,
+ "loss": 2.0883,
+ "step": 138
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016117318435754192,
+ "loss": 2.0803,
+ "step": 139
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016089385474860336,
+ "loss": 2.0617,
+ "step": 140
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016061452513966483,
+ "loss": 2.1265,
+ "step": 141
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016033519553072627,
+ "loss": 2.0151,
+ "step": 142
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0001600558659217877,
+ "loss": 1.996,
+ "step": 143
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.00015977653631284918,
+ "loss": 2.0164,
+ "step": 144
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.00015949720670391061,
+ "loss": 2.0314,
+ "step": 145
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015921787709497208,
+ "loss": 1.9501,
+ "step": 146
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015893854748603352,
+ "loss": 2.087,
+ "step": 147
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.000158659217877095,
+ "loss": 2.0262,
+ "step": 148
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.00015837988826815643,
+ "loss": 2.0765,
+ "step": 149
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0001581005586592179,
+ "loss": 2.105,
+ "step": 150
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.00015782122905027933,
+ "loss": 1.9863,
+ "step": 151
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0001575418994413408,
+ "loss": 1.9873,
+ "step": 152
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.00015726256983240224,
+ "loss": 2.0094,
+ "step": 153
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.00015698324022346368,
+ "loss": 1.9141,
+ "step": 154
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015670391061452514,
+ "loss": 1.917,
+ "step": 155
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015642458100558658,
+ "loss": 2.109,
+ "step": 156
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015614525139664805,
+ "loss": 1.9799,
+ "step": 157
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015586592178770952,
+ "loss": 1.9571,
+ "step": 158
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015558659217877096,
+ "loss": 1.9931,
+ "step": 159
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015530726256983242,
+ "loss": 2.1004,
+ "step": 160
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.00015502793296089386,
+ "loss": 2.0385,
+ "step": 161
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0001547486033519553,
+ "loss": 1.9751,
+ "step": 162
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.00015446927374301677,
+ "loss": 2.0544,
+ "step": 163
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0001541899441340782,
+ "loss": 2.0069,
+ "step": 164
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.00015391061452513967,
+ "loss": 1.9576,
+ "step": 165
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015363128491620114,
+ "loss": 1.8991,
+ "step": 166
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015335195530726258,
+ "loss": 1.9336,
+ "step": 167
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015307262569832405,
+ "loss": 1.9736,
+ "step": 168
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015279329608938548,
+ "loss": 1.9702,
+ "step": 169
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.00015251396648044692,
+ "loss": 1.9055,
+ "step": 170
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0001522346368715084,
+ "loss": 2.0503,
+ "step": 171
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.00015195530726256983,
+ "loss": 2.0039,
+ "step": 172
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0001516759776536313,
+ "loss": 1.9406,
+ "step": 173
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.00015139664804469274,
+ "loss": 2.0525,
+ "step": 174
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0001511173184357542,
+ "loss": 1.9234,
+ "step": 175
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.00015083798882681567,
+ "loss": 1.8614,
+ "step": 176
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0001505586592178771,
+ "loss": 1.9616,
+ "step": 177
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.00015027932960893855,
+ "loss": 1.9509,
+ "step": 178
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.9592,
+ "step": 179
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014972067039106145,
+ "loss": 1.8991,
+ "step": 180
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014944134078212292,
+ "loss": 1.9127,
+ "step": 181
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.00014916201117318436,
+ "loss": 1.8982,
+ "step": 182
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0001488826815642458,
+ "loss": 1.9534,
+ "step": 183
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0001486033519553073,
+ "loss": 1.7794,
+ "step": 184
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.00014832402234636873,
+ "loss": 1.7958,
+ "step": 185
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014804469273743017,
+ "loss": 1.8282,
+ "step": 186
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014776536312849164,
+ "loss": 2.0423,
+ "step": 187
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.00014748603351955308,
+ "loss": 1.9282,
+ "step": 188
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014720670391061454,
+ "loss": 1.9072,
+ "step": 189
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014692737430167598,
+ "loss": 1.8665,
+ "step": 190
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.00014664804469273742,
+ "loss": 1.9021,
+ "step": 191
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.0001463687150837989,
+ "loss": 1.7308,
+ "step": 192
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.00014608938547486035,
+ "loss": 1.9165,
+ "step": 193
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.0001458100558659218,
+ "loss": 1.842,
+ "step": 194
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.00014553072625698326,
+ "loss": 1.9128,
+ "step": 195
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.0001452513966480447,
+ "loss": 1.8005,
+ "step": 196
+ },
+ {
+ "epoch": 1.1,
+ "learning_rate": 0.00014497206703910614,
+ "loss": 1.8547,
+ "step": 197
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.0001446927374301676,
+ "loss": 1.9042,
+ "step": 198
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.00014441340782122905,
+ "loss": 1.8609,
+ "step": 199
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.0001441340782122905,
+ "loss": 1.9591,
+ "step": 200
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.00014385474860335195,
+ "loss": 1.8722,
+ "step": 201
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014357541899441342,
+ "loss": 1.8535,
+ "step": 202
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014329608938547488,
+ "loss": 1.8676,
+ "step": 203
+ },
+ {
+ "epoch": 1.14,
+ "learning_rate": 0.00014301675977653632,
+ "loss": 1.8976,
+ "step": 204
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014273743016759776,
+ "loss": 1.7723,
+ "step": 205
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014245810055865923,
+ "loss": 1.8165,
+ "step": 206
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014217877094972067,
+ "loss": 1.7811,
+ "step": 207
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014189944134078214,
+ "loss": 1.908,
+ "step": 208
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014162011173184357,
+ "loss": 1.7663,
+ "step": 209
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014134078212290501,
+ "loss": 1.6779,
+ "step": 210
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.0001410614525139665,
+ "loss": 1.9039,
+ "step": 211
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.00014078212290502795,
+ "loss": 1.8033,
+ "step": 212
+ },
+ {
+ "epoch": 1.19,
+ "learning_rate": 0.00014050279329608939,
+ "loss": 1.8251,
+ "step": 213
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.00014022346368715085,
+ "loss": 1.8505,
+ "step": 214
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.0001399441340782123,
+ "loss": 1.8147,
+ "step": 215
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.00013966480446927376,
+ "loss": 1.8419,
+ "step": 216
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.0001393854748603352,
+ "loss": 1.8401,
+ "step": 217
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.00013910614525139664,
+ "loss": 1.8912,
+ "step": 218
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.0001388268156424581,
+ "loss": 1.7548,
+ "step": 219
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.00013854748603351957,
+ "loss": 1.8741,
+ "step": 220
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.000138268156424581,
+ "loss": 1.9549,
+ "step": 221
+ },
+ {
+ "epoch": 1.24,
+ "learning_rate": 0.00013798882681564248,
+ "loss": 1.9093,
+ "step": 222
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013770949720670392,
+ "loss": 1.7896,
+ "step": 223
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013743016759776538,
+ "loss": 1.8491,
+ "step": 224
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013715083798882682,
+ "loss": 1.7851,
+ "step": 225
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013687150837988826,
+ "loss": 1.6992,
+ "step": 226
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013659217877094973,
+ "loss": 1.9765,
+ "step": 227
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013631284916201117,
+ "loss": 1.8179,
+ "step": 228
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.00013603351955307263,
+ "loss": 1.8548,
+ "step": 229
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.0001357541899441341,
+ "loss": 1.8843,
+ "step": 230
+ },
+ {
+ "epoch": 1.29,
+ "learning_rate": 0.00013547486033519554,
+ "loss": 1.9105,
+ "step": 231
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013519553072625698,
+ "loss": 1.8748,
+ "step": 232
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013491620111731844,
+ "loss": 1.7976,
+ "step": 233
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013463687150837988,
+ "loss": 1.7369,
+ "step": 234
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013435754189944135,
+ "loss": 1.7808,
+ "step": 235
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.0001340782122905028,
+ "loss": 1.8385,
+ "step": 236
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.00013379888268156423,
+ "loss": 1.8295,
+ "step": 237
+ },
+ {
+ "epoch": 1.33,
+ "learning_rate": 0.00013351955307262572,
+ "loss": 1.757,
+ "step": 238
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.00013324022346368716,
+ "loss": 1.7904,
+ "step": 239
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.0001329608938547486,
+ "loss": 1.7632,
+ "step": 240
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.00013268156424581007,
+ "loss": 1.7867,
+ "step": 241
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.0001324022346368715,
+ "loss": 1.8259,
+ "step": 242
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.00013212290502793297,
+ "loss": 1.6655,
+ "step": 243
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.0001318435754189944,
+ "loss": 1.6848,
+ "step": 244
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013156424581005585,
+ "loss": 1.7931,
+ "step": 245
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013128491620111732,
+ "loss": 1.7868,
+ "step": 246
+ },
+ {
+ "epoch": 1.38,
+ "learning_rate": 0.00013100558659217879,
+ "loss": 1.7732,
+ "step": 247
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.00013072625698324022,
+ "loss": 1.7851,
+ "step": 248
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.0001304469273743017,
+ "loss": 1.7406,
+ "step": 249
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.00013016759776536313,
+ "loss": 1.5853,
+ "step": 250
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.0001298882681564246,
+ "loss": 1.8271,
+ "step": 251
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012960893854748604,
+ "loss": 1.6054,
+ "step": 252
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012932960893854748,
+ "loss": 1.6884,
+ "step": 253
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012905027932960894,
+ "loss": 1.7333,
+ "step": 254
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012877094972067038,
+ "loss": 1.803,
+ "step": 255
+ },
+ {
+ "epoch": 1.43,
+ "learning_rate": 0.00012849162011173185,
+ "loss": 1.63,
+ "step": 256
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012821229050279331,
+ "loss": 1.8023,
+ "step": 257
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012793296089385475,
+ "loss": 1.762,
+ "step": 258
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012765363128491622,
+ "loss": 1.7565,
+ "step": 259
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012737430167597766,
+ "loss": 1.5937,
+ "step": 260
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.0001270949720670391,
+ "loss": 1.7208,
+ "step": 261
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.00012681564245810057,
+ "loss": 1.8097,
+ "step": 262
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.000126536312849162,
+ "loss": 1.7042,
+ "step": 263
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.00012625698324022347,
+ "loss": 1.7892,
+ "step": 264
+ },
+ {
+ "epoch": 1.48,
+ "learning_rate": 0.00012597765363128494,
+ "loss": 1.752,
+ "step": 265
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012569832402234638,
+ "loss": 1.7333,
+ "step": 266
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012541899441340784,
+ "loss": 1.7492,
+ "step": 267
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012513966480446928,
+ "loss": 1.818,
+ "step": 268
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012486033519553072,
+ "loss": 1.7146,
+ "step": 269
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.0001245810055865922,
+ "loss": 1.7958,
+ "step": 270
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.00012430167597765363,
+ "loss": 1.7212,
+ "step": 271
+ },
+ {
+ "epoch": 1.52,
+ "learning_rate": 0.0001240223463687151,
+ "loss": 1.8835,
+ "step": 272
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.00012374301675977656,
+ "loss": 1.7486,
+ "step": 273
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.000123463687150838,
+ "loss": 1.8103,
+ "step": 274
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.00012318435754189944,
+ "loss": 1.7557,
+ "step": 275
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.0001229050279329609,
+ "loss": 1.6198,
+ "step": 276
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.00012262569832402235,
+ "loss": 1.6971,
+ "step": 277
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.0001223463687150838,
+ "loss": 1.668,
+ "step": 278
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.00012206703910614525,
+ "loss": 1.8795,
+ "step": 279
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.0001217877094972067,
+ "loss": 1.6412,
+ "step": 280
+ },
+ {
+ "epoch": 1.57,
+ "learning_rate": 0.00012150837988826816,
+ "loss": 1.7497,
+ "step": 281
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012122905027932962,
+ "loss": 1.5577,
+ "step": 282
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012094972067039108,
+ "loss": 1.8049,
+ "step": 283
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012067039106145253,
+ "loss": 1.6834,
+ "step": 284
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012039106145251397,
+ "loss": 1.7978,
+ "step": 285
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00012011173184357542,
+ "loss": 1.6558,
+ "step": 286
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00011983240223463687,
+ "loss": 1.7561,
+ "step": 287
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011955307262569833,
+ "loss": 1.6807,
+ "step": 288
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011927374301675978,
+ "loss": 1.6937,
+ "step": 289
+ },
+ {
+ "epoch": 1.62,
+ "learning_rate": 0.00011899441340782122,
+ "loss": 1.6591,
+ "step": 290
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.0001187150837988827,
+ "loss": 1.6771,
+ "step": 291
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.00011843575418994415,
+ "loss": 1.7743,
+ "step": 292
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011815642458100559,
+ "loss": 1.5857,
+ "step": 293
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011787709497206705,
+ "loss": 1.6999,
+ "step": 294
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.0001175977653631285,
+ "loss": 1.5661,
+ "step": 295
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.00011731843575418995,
+ "loss": 1.7235,
+ "step": 296
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.0001170391061452514,
+ "loss": 1.607,
+ "step": 297
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.00011675977653631284,
+ "loss": 1.68,
+ "step": 298
+ },
+ {
+ "epoch": 1.67,
+ "learning_rate": 0.0001164804469273743,
+ "loss": 1.6938,
+ "step": 299
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011620111731843578,
+ "loss": 1.6315,
+ "step": 300
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011592178770949722,
+ "loss": 1.6802,
+ "step": 301
+ },
+ {
+ "epoch": 1.69,
+ "learning_rate": 0.00011564245810055867,
+ "loss": 1.7174,
+ "step": 302
+ },
+ {
+ "epoch": 1.69,
+ "learning_rate": 0.00011536312849162012,
+ "loss": 1.5212,
+ "step": 303
+ },
+ {
+ "epoch": 1.7,
+ "learning_rate": 0.00011508379888268157,
+ "loss": 1.5808,
+ "step": 304
+ },
+ {
+ "epoch": 1.7,
+ "learning_rate": 0.00011480446927374303,
+ "loss": 1.6152,
+ "step": 305
+ },
+ {
+ "epoch": 1.71,
+ "learning_rate": 0.00011452513966480447,
+ "loss": 1.5435,
+ "step": 306
+ },
+ {
+ "epoch": 1.72,
+ "learning_rate": 0.00011424581005586592,
+ "loss": 1.6603,
+ "step": 307
+ },
+ {
+ "epoch": 1.72,
+ "learning_rate": 0.00011396648044692737,
+ "loss": 1.685,
+ "step": 308
+ },
+ {
+ "epoch": 1.73,
+ "learning_rate": 0.00011368715083798884,
+ "loss": 1.6002,
+ "step": 309
+ },
+ {
+ "epoch": 1.73,
+ "learning_rate": 0.00011340782122905029,
+ "loss": 1.6046,
+ "step": 310
+ },
+ {
+ "epoch": 1.74,
+ "learning_rate": 0.00011312849162011174,
+ "loss": 1.5969,
+ "step": 311
+ },
+ {
+ "epoch": 1.74,
+ "learning_rate": 0.0001128491620111732,
+ "loss": 1.5845,
+ "step": 312
+ },
+ {
+ "epoch": 1.75,
+ "learning_rate": 0.00011256983240223464,
+ "loss": 1.8183,
+ "step": 313
+ },
+ {
+ "epoch": 1.75,
+ "learning_rate": 0.00011229050279329609,
+ "loss": 1.6953,
+ "step": 314
+ },
+ {
+ "epoch": 1.76,
+ "learning_rate": 0.00011201117318435754,
+ "loss": 1.7787,
+ "step": 315
+ },
+ {
+ "epoch": 1.77,
+ "learning_rate": 0.000111731843575419,
+ "loss": 1.6422,
+ "step": 316
+ },
+ {
+ "epoch": 1.77,
+ "learning_rate": 0.00011145251396648045,
+ "loss": 1.7034,
+ "step": 317
+ },
+ {
+ "epoch": 1.78,
+ "learning_rate": 0.00011117318435754192,
+ "loss": 1.7301,
+ "step": 318
+ },
+ {
+ "epoch": 1.78,
+ "learning_rate": 0.00011089385474860337,
+ "loss": 1.7084,
+ "step": 319
+ },
+ {
+ "epoch": 1.79,
+ "learning_rate": 0.00011061452513966482,
+ "loss": 1.772,
+ "step": 320
+ },
+ {
+ "epoch": 1.79,
+ "learning_rate": 0.00011033519553072626,
+ "loss": 1.5733,
+ "step": 321
+ },
+ {
+ "epoch": 1.8,
+ "learning_rate": 0.00011005586592178771,
+ "loss": 1.6423,
+ "step": 322
+ },
+ {
+ "epoch": 1.8,
+ "learning_rate": 0.00010977653631284917,
+ "loss": 1.5809,
+ "step": 323
+ },
+ {
+ "epoch": 1.81,
+ "learning_rate": 0.00010949720670391062,
+ "loss": 1.6781,
+ "step": 324
+ },
+ {
+ "epoch": 1.82,
+ "learning_rate": 0.00010921787709497207,
+ "loss": 1.6788,
+ "step": 325
+ },
+ {
+ "epoch": 1.82,
+ "learning_rate": 0.00010893854748603351,
+ "loss": 1.6346,
+ "step": 326
+ },
+ {
+ "epoch": 1.83,
+ "learning_rate": 0.00010865921787709499,
+ "loss": 1.6634,
+ "step": 327
+ },
+ {
+ "epoch": 1.83,
+ "learning_rate": 0.00010837988826815643,
+ "loss": 1.7561,
+ "step": 328
+ },
+ {
+ "epoch": 1.84,
+ "learning_rate": 0.00010810055865921788,
+ "loss": 1.66,
+ "step": 329
+ },
+ {
+ "epoch": 1.84,
+ "learning_rate": 0.00010782122905027934,
+ "loss": 1.7298,
+ "step": 330
+ },
+ {
+ "epoch": 1.85,
+ "learning_rate": 0.00010754189944134079,
+ "loss": 1.6893,
+ "step": 331
+ },
+ {
+ "epoch": 1.85,
+ "learning_rate": 0.00010726256983240224,
+ "loss": 1.7631,
+ "step": 332
+ },
+ {
+ "epoch": 1.86,
+ "learning_rate": 0.00010698324022346368,
+ "loss": 1.6633,
+ "step": 333
+ },
+ {
+ "epoch": 1.87,
+ "learning_rate": 0.00010670391061452513,
+ "loss": 1.5388,
+ "step": 334
+ },
+ {
+ "epoch": 1.87,
+ "learning_rate": 0.00010642458100558659,
+ "loss": 1.6718,
+ "step": 335
+ },
+ {
+ "epoch": 1.88,
+ "learning_rate": 0.00010614525139664805,
+ "loss": 1.5536,
+ "step": 336
+ },
+ {
+ "epoch": 1.88,
+ "learning_rate": 0.00010586592178770951,
+ "loss": 1.6483,
+ "step": 337
+ },
+ {
+ "epoch": 1.89,
+ "learning_rate": 0.00010558659217877096,
+ "loss": 1.5774,
+ "step": 338
+ },
+ {
+ "epoch": 1.89,
+ "learning_rate": 0.00010530726256983241,
+ "loss": 1.6366,
+ "step": 339
+ },
+ {
+ "epoch": 1.9,
+ "learning_rate": 0.00010502793296089387,
+ "loss": 1.5567,
+ "step": 340
+ },
+ {
+ "epoch": 1.91,
+ "learning_rate": 0.0001047486033519553,
+ "loss": 1.5323,
+ "step": 341
+ },
+ {
+ "epoch": 1.91,
+ "learning_rate": 0.00010446927374301676,
+ "loss": 1.4608,
+ "step": 342
+ },
+ {
+ "epoch": 1.92,
+ "learning_rate": 0.00010418994413407821,
+ "loss": 1.5933,
+ "step": 343
+ },
+ {
+ "epoch": 1.92,
+ "learning_rate": 0.00010391061452513966,
+ "loss": 1.6625,
+ "step": 344
+ },
+ {
+ "epoch": 1.93,
+ "learning_rate": 0.00010363128491620113,
+ "loss": 1.7236,
+ "step": 345
+ },
+ {
+ "epoch": 1.93,
+ "learning_rate": 0.00010335195530726258,
+ "loss": 1.759,
+ "step": 346
+ },
+ {
+ "epoch": 1.94,
+ "learning_rate": 0.00010307262569832404,
+ "loss": 1.7248,
+ "step": 347
+ },
+ {
+ "epoch": 1.94,
+ "learning_rate": 0.00010279329608938548,
+ "loss": 1.5144,
+ "step": 348
+ },
+ {
+ "epoch": 1.95,
+ "learning_rate": 0.00010251396648044693,
+ "loss": 1.6905,
+ "step": 349
+ },
+ {
+ "epoch": 1.96,
+ "learning_rate": 0.00010223463687150838,
+ "loss": 1.6119,
+ "step": 350
+ },
+ {
+ "epoch": 1.96,
+ "learning_rate": 0.00010195530726256983,
+ "loss": 1.5464,
+ "step": 351
+ },
+ {
+ "epoch": 1.97,
+ "learning_rate": 0.00010167597765363129,
+ "loss": 1.6901,
+ "step": 352
+ },
+ {
+ "epoch": 1.97,
+ "learning_rate": 0.00010139664804469273,
+ "loss": 1.3511,
+ "step": 353
+ },
+ {
+ "epoch": 1.98,
+ "learning_rate": 0.0001011173184357542,
+ "loss": 1.5434,
+ "step": 354
+ },
+ {
+ "epoch": 1.98,
+ "learning_rate": 0.00010083798882681566,
+ "loss": 1.5891,
+ "step": 355
+ },
+ {
+ "epoch": 1.99,
+ "learning_rate": 0.0001005586592178771,
+ "loss": 1.6658,
+ "step": 356
+ },
+ {
+ "epoch": 1.99,
+ "learning_rate": 0.00010027932960893855,
+ "loss": 1.5657,
+ "step": 357
+ },
+ {
+ "epoch": 2.0,
+ "learning_rate": 0.0001,
+ "loss": 1.7005,
+ "step": 358
+ },
+ {
+ "epoch": 2.01,
+ "learning_rate": 9.972067039106146e-05,
+ "loss": 1.4202,
+ "step": 359
+ },
+ {
+ "epoch": 2.01,
+ "learning_rate": 9.944134078212291e-05,
+ "loss": 1.5262,
+ "step": 360
+ },
+ {
+ "epoch": 2.02,
+ "learning_rate": 9.916201117318436e-05,
+ "loss": 1.6323,
+ "step": 361
+ },
+ {
+ "epoch": 2.02,
+ "learning_rate": 9.888268156424582e-05,
+ "loss": 1.5521,
+ "step": 362
+ },
+ {
+ "epoch": 2.03,
+ "learning_rate": 9.860335195530727e-05,
+ "loss": 1.5762,
+ "step": 363
+ },
+ {
+ "epoch": 2.03,
+ "learning_rate": 9.832402234636872e-05,
+ "loss": 1.613,
+ "step": 364
+ },
+ {
+ "epoch": 2.04,
+ "learning_rate": 9.804469273743018e-05,
+ "loss": 1.4231,
+ "step": 365
+ },
+ {
+ "epoch": 2.04,
+ "learning_rate": 9.776536312849163e-05,
+ "loss": 1.5706,
+ "step": 366
+ },
+ {
+ "epoch": 2.05,
+ "learning_rate": 9.748603351955308e-05,
+ "loss": 1.5245,
+ "step": 367
+ },
+ {
+ "epoch": 2.06,
+ "learning_rate": 9.720670391061453e-05,
+ "loss": 1.4771,
+ "step": 368
+ },
+ {
+ "epoch": 2.06,
+ "learning_rate": 9.692737430167597e-05,
+ "loss": 1.596,
+ "step": 369
+ },
+ {
+ "epoch": 2.07,
+ "learning_rate": 9.664804469273744e-05,
+ "loss": 1.537,
+ "step": 370
+ },
+ {
+ "epoch": 2.07,
+ "learning_rate": 9.636871508379889e-05,
+ "loss": 1.4276,
+ "step": 371
+ },
+ {
+ "epoch": 2.08,
+ "learning_rate": 9.608938547486033e-05,
+ "loss": 1.4746,
+ "step": 372
+ },
+ {
+ "epoch": 2.08,
+ "learning_rate": 9.581005586592178e-05,
+ "loss": 1.4374,
+ "step": 373
+ },
+ {
+ "epoch": 2.09,
+ "learning_rate": 9.553072625698325e-05,
+ "loss": 1.4704,
+ "step": 374
+ },
+ {
+ "epoch": 2.09,
+ "learning_rate": 9.52513966480447e-05,
+ "loss": 1.5997,
+ "step": 375
+ },
+ {
+ "epoch": 2.1,
+ "learning_rate": 9.497206703910614e-05,
+ "loss": 1.5034,
+ "step": 376
+ },
+ {
+ "epoch": 2.11,
+ "learning_rate": 9.46927374301676e-05,
+ "loss": 1.6392,
+ "step": 377
+ },
+ {
+ "epoch": 2.11,
+ "learning_rate": 9.441340782122905e-05,
+ "loss": 1.5611,
+ "step": 378
+ },
+ {
+ "epoch": 2.12,
+ "learning_rate": 9.413407821229052e-05,
+ "loss": 1.3384,
+ "step": 379
+ },
+ {
+ "epoch": 2.12,
+ "learning_rate": 9.385474860335196e-05,
+ "loss": 1.5732,
+ "step": 380
+ },
+ {
+ "epoch": 2.13,
+ "learning_rate": 9.357541899441341e-05,
+ "loss": 1.3874,
+ "step": 381
+ },
+ {
+ "epoch": 2.13,
+ "learning_rate": 9.329608938547486e-05,
+ "loss": 1.4139,
+ "step": 382
+ },
+ {
+ "epoch": 2.14,
+ "learning_rate": 9.301675977653633e-05,
+ "loss": 1.5195,
+ "step": 383
+ },
+ {
+ "epoch": 2.15,
+ "learning_rate": 9.273743016759777e-05,
+ "loss": 1.4371,
+ "step": 384
+ },
+ {
+ "epoch": 2.15,
+ "learning_rate": 9.245810055865922e-05,
+ "loss": 1.4411,
+ "step": 385
+ },
+ {
+ "epoch": 2.16,
+ "learning_rate": 9.217877094972067e-05,
+ "loss": 1.5214,
+ "step": 386
+ },
+ {
+ "epoch": 2.16,
+ "learning_rate": 9.189944134078213e-05,
+ "loss": 1.4971,
+ "step": 387
+ },
+ {
+ "epoch": 2.17,
+ "learning_rate": 9.162011173184358e-05,
+ "loss": 1.3453,
+ "step": 388
+ },
+ {
+ "epoch": 2.17,
+ "learning_rate": 9.134078212290503e-05,
+ "loss": 1.394,
+ "step": 389
+ },
+ {
+ "epoch": 2.18,
+ "learning_rate": 9.106145251396648e-05,
+ "loss": 1.5058,
+ "step": 390
+ },
+ {
+ "epoch": 2.18,
+ "learning_rate": 9.078212290502794e-05,
+ "loss": 1.4855,
+ "step": 391
+ },
+ {
+ "epoch": 2.19,
+ "learning_rate": 9.050279329608939e-05,
+ "loss": 1.4647,
+ "step": 392
+ },
+ {
+ "epoch": 2.2,
+ "learning_rate": 9.022346368715084e-05,
+ "loss": 1.3435,
+ "step": 393
+ },
+ {
+ "epoch": 2.2,
+ "learning_rate": 8.99441340782123e-05,
+ "loss": 1.5815,
+ "step": 394
+ },
+ {
+ "epoch": 2.21,
+ "learning_rate": 8.966480446927375e-05,
+ "loss": 1.4742,
+ "step": 395
+ },
+ {
+ "epoch": 2.21,
+ "learning_rate": 8.938547486033519e-05,
+ "loss": 1.5389,
+ "step": 396
+ },
+ {
+ "epoch": 2.22,
+ "learning_rate": 8.910614525139666e-05,
+ "loss": 1.384,
+ "step": 397
+ },
+ {
+ "epoch": 2.22,
+ "learning_rate": 8.882681564245811e-05,
+ "loss": 1.3967,
+ "step": 398
+ },
+ {
+ "epoch": 2.23,
+ "learning_rate": 8.854748603351956e-05,
+ "loss": 1.442,
+ "step": 399
+ },
+ {
+ "epoch": 2.23,
+ "learning_rate": 8.8268156424581e-05,
+ "loss": 1.396,
+ "step": 400
+ },
+ {
+ "epoch": 2.24,
+ "learning_rate": 8.798882681564247e-05,
+ "loss": 1.4706,
+ "step": 401
+ },
+ {
+ "epoch": 2.25,
+ "learning_rate": 8.770949720670392e-05,
+ "loss": 1.4217,
+ "step": 402
+ },
+ {
+ "epoch": 2.25,
+ "learning_rate": 8.743016759776537e-05,
+ "loss": 1.5617,
+ "step": 403
+ },
+ {
+ "epoch": 2.26,
+ "learning_rate": 8.715083798882681e-05,
+ "loss": 1.5701,
+ "step": 404
+ },
+ {
+ "epoch": 2.26,
+ "learning_rate": 8.687150837988828e-05,
+ "loss": 1.3659,
+ "step": 405
+ },
+ {
+ "epoch": 2.27,
+ "learning_rate": 8.659217877094973e-05,
+ "loss": 1.3424,
+ "step": 406
+ },
+ {
+ "epoch": 2.27,
+ "learning_rate": 8.631284916201118e-05,
+ "loss": 1.5782,
+ "step": 407
+ },
+ {
+ "epoch": 2.28,
+ "learning_rate": 8.603351955307262e-05,
+ "loss": 1.3355,
+ "step": 408
+ },
+ {
+ "epoch": 2.28,
+ "learning_rate": 8.575418994413408e-05,
+ "loss": 1.4798,
+ "step": 409
+ },
+ {
+ "epoch": 2.29,
+ "learning_rate": 8.547486033519554e-05,
+ "loss": 1.3813,
+ "step": 410
+ },
+ {
+ "epoch": 2.3,
+ "learning_rate": 8.519553072625698e-05,
+ "loss": 1.6011,
+ "step": 411
+ },
+ {
+ "epoch": 2.3,
+ "learning_rate": 8.491620111731844e-05,
+ "loss": 1.4596,
+ "step": 412
+ },
+ {
+ "epoch": 2.31,
+ "learning_rate": 8.463687150837989e-05,
+ "loss": 1.4997,
+ "step": 413
+ },
+ {
+ "epoch": 2.31,
+ "learning_rate": 8.435754189944135e-05,
+ "loss": 1.5255,
+ "step": 414
+ },
+ {
+ "epoch": 2.32,
+ "learning_rate": 8.40782122905028e-05,
+ "loss": 1.5305,
+ "step": 415
+ },
+ {
+ "epoch": 2.32,
+ "learning_rate": 8.379888268156425e-05,
+ "loss": 1.3029,
+ "step": 416
+ },
+ {
+ "epoch": 2.33,
+ "learning_rate": 8.35195530726257e-05,
+ "loss": 1.2046,
+ "step": 417
+ },
+ {
+ "epoch": 2.34,
+ "learning_rate": 8.324022346368715e-05,
+ "loss": 1.4985,
+ "step": 418
+ },
+ {
+ "epoch": 2.34,
+ "learning_rate": 8.29608938547486e-05,
+ "loss": 1.3486,
+ "step": 419
+ },
+ {
+ "epoch": 2.35,
+ "learning_rate": 8.268156424581006e-05,
+ "loss": 1.2839,
+ "step": 420
+ },
+ {
+ "epoch": 2.35,
+ "learning_rate": 8.240223463687151e-05,
+ "loss": 1.4286,
+ "step": 421
+ },
+ {
+ "epoch": 2.36,
+ "learning_rate": 8.212290502793296e-05,
+ "loss": 1.3893,
+ "step": 422
+ },
+ {
+ "epoch": 2.36,
+ "learning_rate": 8.184357541899442e-05,
+ "loss": 1.5477,
+ "step": 423
+ },
+ {
+ "epoch": 2.37,
+ "learning_rate": 8.156424581005587e-05,
+ "loss": 1.4973,
+ "step": 424
+ },
+ {
+ "epoch": 2.37,
+ "learning_rate": 8.128491620111732e-05,
+ "loss": 1.3549,
+ "step": 425
+ },
+ {
+ "epoch": 2.38,
+ "learning_rate": 8.100558659217878e-05,
+ "loss": 1.3966,
+ "step": 426
+ },
+ {
+ "epoch": 2.39,
+ "learning_rate": 8.072625698324023e-05,
+ "loss": 1.3974,
+ "step": 427
+ },
+ {
+ "epoch": 2.39,
+ "learning_rate": 8.044692737430168e-05,
+ "loss": 1.5211,
+ "step": 428
+ },
+ {
+ "epoch": 2.4,
+ "learning_rate": 8.016759776536313e-05,
+ "loss": 1.4751,
+ "step": 429
+ },
+ {
+ "epoch": 2.4,
+ "learning_rate": 7.988826815642459e-05,
+ "loss": 1.4322,
+ "step": 430
+ },
+ {
+ "epoch": 2.41,
+ "learning_rate": 7.960893854748604e-05,
+ "loss": 1.5991,
+ "step": 431
+ },
+ {
+ "epoch": 2.41,
+ "learning_rate": 7.93296089385475e-05,
+ "loss": 1.3288,
+ "step": 432
+ },
+ {
+ "epoch": 2.42,
+ "learning_rate": 7.905027932960895e-05,
+ "loss": 1.4098,
+ "step": 433
+ },
+ {
+ "epoch": 2.42,
+ "learning_rate": 7.87709497206704e-05,
+ "loss": 1.4819,
+ "step": 434
+ },
+ {
+ "epoch": 2.43,
+ "learning_rate": 7.849162011173184e-05,
+ "loss": 1.4003,
+ "step": 435
+ },
+ {
+ "epoch": 2.44,
+ "learning_rate": 7.821229050279329e-05,
+ "loss": 1.3475,
+ "step": 436
+ },
+ {
+ "epoch": 2.44,
+ "learning_rate": 7.793296089385476e-05,
+ "loss": 1.3354,
+ "step": 437
+ },
+ {
+ "epoch": 2.45,
+ "learning_rate": 7.765363128491621e-05,
+ "loss": 1.3144,
+ "step": 438
+ },
+ {
+ "epoch": 2.45,
+ "learning_rate": 7.737430167597765e-05,
+ "loss": 1.4118,
+ "step": 439
+ },
+ {
+ "epoch": 2.46,
+ "learning_rate": 7.70949720670391e-05,
+ "loss": 1.3716,
+ "step": 440
+ },
+ {
+ "epoch": 2.46,
+ "learning_rate": 7.681564245810057e-05,
+ "loss": 1.4191,
+ "step": 441
+ },
+ {
+ "epoch": 2.47,
+ "learning_rate": 7.653631284916202e-05,
+ "loss": 1.5173,
+ "step": 442
+ },
+ {
+ "epoch": 2.47,
+ "learning_rate": 7.625698324022346e-05,
+ "loss": 1.3005,
+ "step": 443
+ },
+ {
+ "epoch": 2.48,
+ "learning_rate": 7.597765363128491e-05,
+ "loss": 1.4586,
+ "step": 444
+ },
+ {
+ "epoch": 2.49,
+ "learning_rate": 7.569832402234637e-05,
+ "loss": 1.5086,
+ "step": 445
+ },
+ {
+ "epoch": 2.49,
+ "learning_rate": 7.541899441340783e-05,
+ "loss": 1.5446,
+ "step": 446
+ },
+ {
+ "epoch": 2.5,
+ "learning_rate": 7.513966480446927e-05,
+ "loss": 1.3274,
+ "step": 447
+ },
+ {
+ "epoch": 2.5,
+ "learning_rate": 7.486033519553073e-05,
+ "loss": 1.4424,
+ "step": 448
+ },
+ {
+ "epoch": 2.51,
+ "learning_rate": 7.458100558659218e-05,
+ "loss": 1.2836,
+ "step": 449
+ },
+ {
+ "epoch": 2.51,
+ "learning_rate": 7.430167597765365e-05,
+ "loss": 1.4322,
+ "step": 450
+ },
+ {
+ "epoch": 2.52,
+ "learning_rate": 7.402234636871509e-05,
+ "loss": 1.4168,
+ "step": 451
+ },
+ {
+ "epoch": 2.53,
+ "learning_rate": 7.374301675977654e-05,
+ "loss": 1.3892,
+ "step": 452
+ },
+ {
+ "epoch": 2.53,
+ "learning_rate": 7.346368715083799e-05,
+ "loss": 1.6538,
+ "step": 453
+ },
+ {
+ "epoch": 2.54,
+ "learning_rate": 7.318435754189944e-05,
+ "loss": 1.4573,
+ "step": 454
+ },
+ {
+ "epoch": 2.54,
+ "learning_rate": 7.29050279329609e-05,
+ "loss": 1.4593,
+ "step": 455
+ },
+ {
+ "epoch": 2.55,
+ "learning_rate": 7.262569832402235e-05,
+ "loss": 1.4316,
+ "step": 456
+ },
+ {
+ "epoch": 2.55,
+ "learning_rate": 7.23463687150838e-05,
+ "loss": 1.2702,
+ "step": 457
+ },
+ {
+ "epoch": 2.56,
+ "learning_rate": 7.206703910614526e-05,
+ "loss": 1.5458,
+ "step": 458
+ },
+ {
+ "epoch": 2.56,
+ "learning_rate": 7.178770949720671e-05,
+ "loss": 1.5179,
+ "step": 459
+ },
+ {
+ "epoch": 2.57,
+ "learning_rate": 7.150837988826816e-05,
+ "loss": 1.3844,
+ "step": 460
+ },
+ {
+ "epoch": 2.58,
+ "learning_rate": 7.122905027932961e-05,
+ "loss": 1.3698,
+ "step": 461
+ },
+ {
+ "epoch": 2.58,
+ "learning_rate": 7.094972067039107e-05,
+ "loss": 1.4396,
+ "step": 462
+ },
+ {
+ "epoch": 2.59,
+ "learning_rate": 7.067039106145251e-05,
+ "loss": 1.5486,
+ "step": 463
+ },
+ {
+ "epoch": 2.59,
+ "learning_rate": 7.039106145251397e-05,
+ "loss": 1.2032,
+ "step": 464
+ },
+ {
+ "epoch": 2.6,
+ "learning_rate": 7.011173184357543e-05,
+ "loss": 1.3434,
+ "step": 465
+ },
+ {
+ "epoch": 2.6,
+ "learning_rate": 6.983240223463688e-05,
+ "loss": 1.4065,
+ "step": 466
+ },
+ {
+ "epoch": 2.61,
+ "learning_rate": 6.955307262569832e-05,
+ "loss": 1.3362,
+ "step": 467
+ },
+ {
+ "epoch": 2.61,
+ "learning_rate": 6.927374301675979e-05,
+ "loss": 1.375,
+ "step": 468
+ },
+ {
+ "epoch": 2.62,
+ "learning_rate": 6.899441340782124e-05,
+ "loss": 1.4549,
+ "step": 469
+ },
+ {
+ "epoch": 2.63,
+ "learning_rate": 6.871508379888269e-05,
+ "loss": 1.3724,
+ "step": 470
+ },
+ {
+ "epoch": 2.63,
+ "learning_rate": 6.843575418994413e-05,
+ "loss": 1.1602,
+ "step": 471
+ },
+ {
+ "epoch": 2.64,
+ "learning_rate": 6.815642458100558e-05,
+ "loss": 1.4346,
+ "step": 472
+ },
+ {
+ "epoch": 2.64,
+ "learning_rate": 6.787709497206705e-05,
+ "loss": 1.2877,
+ "step": 473
+ },
+ {
+ "epoch": 2.65,
+ "learning_rate": 6.759776536312849e-05,
+ "loss": 1.394,
+ "step": 474
+ },
+ {
+ "epoch": 2.65,
+ "learning_rate": 6.731843575418994e-05,
+ "loss": 1.3387,
+ "step": 475
+ },
+ {
+ "epoch": 2.66,
+ "learning_rate": 6.70391061452514e-05,
+ "loss": 1.41,
+ "step": 476
+ },
+ {
+ "epoch": 2.66,
+ "learning_rate": 6.675977653631286e-05,
+ "loss": 1.4299,
+ "step": 477
+ },
+ {
+ "epoch": 2.67,
+ "learning_rate": 6.64804469273743e-05,
+ "loss": 1.3824,
+ "step": 478
+ },
+ {
+ "epoch": 2.68,
+ "learning_rate": 6.620111731843575e-05,
+ "loss": 1.2656,
+ "step": 479
+ },
+ {
+ "epoch": 2.68,
+ "learning_rate": 6.59217877094972e-05,
+ "loss": 1.2811,
+ "step": 480
+ },
+ {
+ "epoch": 2.69,
+ "learning_rate": 6.564245810055866e-05,
+ "loss": 1.6839,
+ "step": 481
+ },
+ {
+ "epoch": 2.69,
+ "learning_rate": 6.536312849162011e-05,
+ "loss": 1.3089,
+ "step": 482
+ },
+ {
+ "epoch": 2.7,
+ "learning_rate": 6.508379888268157e-05,
+ "loss": 1.4323,
+ "step": 483
+ },
+ {
+ "epoch": 2.7,
+ "learning_rate": 6.480446927374302e-05,
+ "loss": 1.3439,
+ "step": 484
+ },
+ {
+ "epoch": 2.71,
+ "learning_rate": 6.452513966480447e-05,
+ "loss": 1.3614,
+ "step": 485
+ },
+ {
+ "epoch": 2.72,
+ "learning_rate": 6.424581005586592e-05,
+ "loss": 1.4974,
+ "step": 486
+ },
+ {
+ "epoch": 2.72,
+ "learning_rate": 6.396648044692738e-05,
+ "loss": 1.4963,
+ "step": 487
+ },
+ {
+ "epoch": 2.73,
+ "learning_rate": 6.368715083798883e-05,
+ "loss": 1.2902,
+ "step": 488
+ },
+ {
+ "epoch": 2.73,
+ "learning_rate": 6.340782122905028e-05,
+ "loss": 1.3226,
+ "step": 489
+ },
+ {
+ "epoch": 2.74,
+ "learning_rate": 6.312849162011174e-05,
+ "loss": 1.4353,
+ "step": 490
+ },
+ {
+ "epoch": 2.74,
+ "learning_rate": 6.284916201117319e-05,
+ "loss": 1.281,
+ "step": 491
+ },
+ {
+ "epoch": 2.75,
+ "learning_rate": 6.256983240223464e-05,
+ "loss": 1.3755,
+ "step": 492
+ },
+ {
+ "epoch": 2.75,
+ "learning_rate": 6.22905027932961e-05,
+ "loss": 1.3811,
+ "step": 493
+ },
+ {
+ "epoch": 2.76,
+ "learning_rate": 6.201117318435755e-05,
+ "loss": 1.4627,
+ "step": 494
+ },
+ {
+ "epoch": 2.77,
+ "learning_rate": 6.1731843575419e-05,
+ "loss": 1.3482,
+ "step": 495
+ },
+ {
+ "epoch": 2.77,
+ "learning_rate": 6.145251396648045e-05,
+ "loss": 1.4274,
+ "step": 496
+ },
+ {
+ "epoch": 2.78,
+ "learning_rate": 6.11731843575419e-05,
+ "loss": 1.3465,
+ "step": 497
+ },
+ {
+ "epoch": 2.78,
+ "learning_rate": 6.089385474860335e-05,
+ "loss": 1.2319,
+ "step": 498
+ },
+ {
+ "epoch": 2.79,
+ "learning_rate": 6.061452513966481e-05,
+ "loss": 1.2835,
+ "step": 499
+ },
+ {
+ "epoch": 2.79,
+ "learning_rate": 6.0335195530726265e-05,
+ "loss": 1.4836,
+ "step": 500
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 716,
+ "num_train_epochs": 4,
+ "save_steps": 100,
+ "total_flos": 2.572486267094016e+17,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-500/training_args.bin b/checkpoint-500/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8672c716e925d0028b4938db147703f58656ff7
--- /dev/null
+++ b/checkpoint-500/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:188ae1c421cc0c6435d1f71d8d3423ac4abc7dba0e6fc2efcbc4dbe77c741317
+size 4027
diff --git a/checkpoint-600/README.md b/checkpoint-600/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..08371015f02382e6fcba318f4aaea54ae52cd3c4
--- /dev/null
+++ b/checkpoint-600/README.md
@@ -0,0 +1,34 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+### Framework versions
+
+- PEFT 0.6.0.dev0
+
+- PEFT 0.6.0.dev0
diff --git a/checkpoint-600/adapter_config.json b/checkpoint-600/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c661f74e40394c079de68f07861d73c341803b67
--- /dev/null
+++ b/checkpoint-600/adapter_config.json
@@ -0,0 +1,25 @@
+{
+ "alpha_pattern": {},
+ "auto_mapping": null,
+ "base_model_name_or_path": "bigscience/bloomz-3b",
+ "bias": "none",
+ "fan_in_fan_out": false,
+ "inference_mode": true,
+ "init_lora_weights": true,
+ "layers_pattern": null,
+ "layers_to_transform": null,
+ "lora_alpha": 16,
+ "lora_dropout": 0.0,
+ "modules_to_save": null,
+ "peft_type": "LORA",
+ "r": 8,
+ "rank_pattern": {},
+ "revision": null,
+ "target_modules": [
+ "dense_4h_to_h",
+ "dense",
+ "dense_h_to_4h",
+ "query_key_value"
+ ],
+ "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/checkpoint-600/adapter_model.bin b/checkpoint-600/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..8fbb1832de137e033b6dbf7ffc70284e5becb659
--- /dev/null
+++ b/checkpoint-600/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:435ffb177831605f5479223de5dfef343a64e6bc9f2d4e05fba2ab86e3d42c9b
+size 39409357
diff --git a/checkpoint-600/optimizer.pt b/checkpoint-600/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..64d125cfb1d1abce95ae8935c5159c5c4afc4d85
--- /dev/null
+++ b/checkpoint-600/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a67c0a0a04a16e60dc6df9ce52ee43c4b51d5e34c198de23f6afed8ff5bd6d3e
+size 78844421
diff --git a/checkpoint-600/rng_state.pth b/checkpoint-600/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..452f99ac534b9117d836494d73222e3d44e1523b
--- /dev/null
+++ b/checkpoint-600/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6869750f95a25c4e970298a33adf90e2d7ab52680bf3317239bff1b10103235
+size 14575
diff --git a/checkpoint-600/scheduler.pt b/checkpoint-600/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..dc91f19f9994d9dc5e2cb196fcfd876e3f7fe0c8
--- /dev/null
+++ b/checkpoint-600/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df189dbec1d9da6456fed4321f8fb7b9b04184c73d2cc55b1372ecebfa9aaaff
+size 627
diff --git a/checkpoint-600/special_tokens_map.json b/checkpoint-600/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdafe480f024ff444c7492147536765ce5d55a2d
--- /dev/null
+++ b/checkpoint-600/special_tokens_map.json
@@ -0,0 +1,6 @@
+{
+ "bos_token": "",
+ "eos_token": "",
+ "pad_token": "",
+ "unk_token": ""
+}
diff --git a/checkpoint-600/tokenizer.json b/checkpoint-600/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..673c31abdeadf6576c3c754df86459e1ad64e207
--- /dev/null
+++ b/checkpoint-600/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85b00d7db4df5df2e3f01cacc3feda246002a672f3356eec7f4b04a22eb0dfbe
+size 14500570
diff --git a/checkpoint-600/tokenizer_config.json b/checkpoint-600/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b56cc9c2965c07132c35df3e2972e93d98c82c3
--- /dev/null
+++ b/checkpoint-600/tokenizer_config.json
@@ -0,0 +1,10 @@
+{
+ "add_prefix_space": false,
+ "bos_token": "",
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "",
+ "model_max_length": 1000000000000000019884624838656,
+ "pad_token": "",
+ "tokenizer_class": "BloomTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoint-600/trainer_state.json b/checkpoint-600/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..779b4f6d8a9d155187db4987b1d9f9ebcb34f156
--- /dev/null
+++ b/checkpoint-600/trainer_state.json
@@ -0,0 +1,3619 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 3.35195530726257,
+ "eval_steps": 500,
+ "global_step": 600,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019972067039106145,
+ "loss": 2.6443,
+ "step": 1
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019944134078212292,
+ "loss": 2.4104,
+ "step": 2
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019916201117318435,
+ "loss": 2.4975,
+ "step": 3
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019888268156424582,
+ "loss": 2.3513,
+ "step": 4
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0001986033519553073,
+ "loss": 2.4274,
+ "step": 5
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.00019832402234636873,
+ "loss": 2.3628,
+ "step": 6
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0001980446927374302,
+ "loss": 2.3567,
+ "step": 7
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.00019776536312849163,
+ "loss": 2.4121,
+ "step": 8
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.00019748603351955307,
+ "loss": 2.4033,
+ "step": 9
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019720670391061454,
+ "loss": 2.2805,
+ "step": 10
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019692737430167598,
+ "loss": 2.2639,
+ "step": 11
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019664804469273744,
+ "loss": 2.2724,
+ "step": 12
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019636871508379888,
+ "loss": 2.332,
+ "step": 13
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019608938547486035,
+ "loss": 2.2261,
+ "step": 14
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019581005586592182,
+ "loss": 2.2208,
+ "step": 15
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.00019553072625698326,
+ "loss": 2.3351,
+ "step": 16
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0001952513966480447,
+ "loss": 2.2475,
+ "step": 17
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.00019497206703910616,
+ "loss": 2.3283,
+ "step": 18
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0001946927374301676,
+ "loss": 2.1346,
+ "step": 19
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.00019441340782122907,
+ "loss": 2.131,
+ "step": 20
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0001941340782122905,
+ "loss": 2.1718,
+ "step": 21
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.00019385474860335195,
+ "loss": 2.2446,
+ "step": 22
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0001935754189944134,
+ "loss": 2.306,
+ "step": 23
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.00019329608938547488,
+ "loss": 2.1908,
+ "step": 24
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.00019301675977653632,
+ "loss": 2.2844,
+ "step": 25
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019273743016759779,
+ "loss": 2.2235,
+ "step": 26
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019245810055865922,
+ "loss": 2.1842,
+ "step": 27
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019217877094972066,
+ "loss": 2.2675,
+ "step": 28
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019189944134078213,
+ "loss": 2.2532,
+ "step": 29
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019162011173184357,
+ "loss": 2.1788,
+ "step": 30
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019134078212290504,
+ "loss": 2.2494,
+ "step": 31
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0001910614525139665,
+ "loss": 2.1995,
+ "step": 32
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.00019078212290502794,
+ "loss": 2.1451,
+ "step": 33
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0001905027932960894,
+ "loss": 2.223,
+ "step": 34
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.00019022346368715085,
+ "loss": 2.2854,
+ "step": 35
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0001899441340782123,
+ "loss": 2.2265,
+ "step": 36
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.00018966480446927375,
+ "loss": 2.1214,
+ "step": 37
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0001893854748603352,
+ "loss": 2.1898,
+ "step": 38
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.00018910614525139666,
+ "loss": 2.1974,
+ "step": 39
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0001888268156424581,
+ "loss": 2.2259,
+ "step": 40
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018854748603351957,
+ "loss": 2.2094,
+ "step": 41
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018826815642458103,
+ "loss": 2.1731,
+ "step": 42
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.00018798882681564247,
+ "loss": 2.2373,
+ "step": 43
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0001877094972067039,
+ "loss": 2.2295,
+ "step": 44
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.00018743016759776538,
+ "loss": 2.1947,
+ "step": 45
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018715083798882682,
+ "loss": 2.2115,
+ "step": 46
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018687150837988828,
+ "loss": 2.1224,
+ "step": 47
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018659217877094972,
+ "loss": 2.2137,
+ "step": 48
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018631284916201116,
+ "loss": 2.2338,
+ "step": 49
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.00018603351955307266,
+ "loss": 2.1298,
+ "step": 50
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0001857541899441341,
+ "loss": 2.0883,
+ "step": 51
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.00018547486033519553,
+ "loss": 2.1216,
+ "step": 52
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.000185195530726257,
+ "loss": 2.2112,
+ "step": 53
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.00018491620111731844,
+ "loss": 2.1224,
+ "step": 54
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0001846368715083799,
+ "loss": 2.2375,
+ "step": 55
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.00018435754189944135,
+ "loss": 2.2235,
+ "step": 56
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018407821229050279,
+ "loss": 2.1682,
+ "step": 57
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018379888268156425,
+ "loss": 2.2077,
+ "step": 58
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.00018351955307262572,
+ "loss": 2.1596,
+ "step": 59
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018324022346368716,
+ "loss": 2.1311,
+ "step": 60
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018296089385474862,
+ "loss": 2.1333,
+ "step": 61
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018268156424581006,
+ "loss": 2.0901,
+ "step": 62
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018240223463687153,
+ "loss": 2.1971,
+ "step": 63
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.00018212290502793297,
+ "loss": 2.2602,
+ "step": 64
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0001818435754189944,
+ "loss": 2.2194,
+ "step": 65
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018156424581005588,
+ "loss": 2.1218,
+ "step": 66
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018128491620111731,
+ "loss": 2.2049,
+ "step": 67
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.00018100558659217878,
+ "loss": 2.1521,
+ "step": 68
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.00018072625698324025,
+ "loss": 2.112,
+ "step": 69
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0001804469273743017,
+ "loss": 2.1906,
+ "step": 70
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.00018016759776536313,
+ "loss": 2.1717,
+ "step": 71
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0001798882681564246,
+ "loss": 2.0712,
+ "step": 72
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.00017960893854748603,
+ "loss": 2.141,
+ "step": 73
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0001793296089385475,
+ "loss": 2.0656,
+ "step": 74
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017905027932960894,
+ "loss": 2.1125,
+ "step": 75
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017877094972067038,
+ "loss": 2.0869,
+ "step": 76
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.00017849162011173187,
+ "loss": 2.2478,
+ "step": 77
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0001782122905027933,
+ "loss": 2.1535,
+ "step": 78
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.00017793296089385475,
+ "loss": 2.1927,
+ "step": 79
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017765363128491622,
+ "loss": 2.1213,
+ "step": 80
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017737430167597766,
+ "loss": 2.0981,
+ "step": 81
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017709497206703912,
+ "loss": 2.1828,
+ "step": 82
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017681564245810056,
+ "loss": 2.0562,
+ "step": 83
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.000176536312849162,
+ "loss": 2.1334,
+ "step": 84
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.00017625698324022347,
+ "loss": 2.1225,
+ "step": 85
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.00017597765363128493,
+ "loss": 2.2098,
+ "step": 86
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017569832402234637,
+ "loss": 2.1519,
+ "step": 87
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017541899441340784,
+ "loss": 2.1132,
+ "step": 88
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017513966480446928,
+ "loss": 2.0333,
+ "step": 89
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017486033519553075,
+ "loss": 2.2764,
+ "step": 90
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017458100558659218,
+ "loss": 2.1838,
+ "step": 91
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017430167597765362,
+ "loss": 2.1386,
+ "step": 92
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0001740223463687151,
+ "loss": 2.1034,
+ "step": 93
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.00017374301675977656,
+ "loss": 2.0346,
+ "step": 94
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.000173463687150838,
+ "loss": 2.0274,
+ "step": 95
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.00017318435754189946,
+ "loss": 2.1036,
+ "step": 96
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0001729050279329609,
+ "loss": 2.1208,
+ "step": 97
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.00017262569832402237,
+ "loss": 2.0572,
+ "step": 98
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0001723463687150838,
+ "loss": 2.1702,
+ "step": 99
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.00017206703910614525,
+ "loss": 2.1302,
+ "step": 100
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0001717877094972067,
+ "loss": 2.0175,
+ "step": 101
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.00017150837988826815,
+ "loss": 2.1006,
+ "step": 102
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017122905027932962,
+ "loss": 2.0662,
+ "step": 103
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017094972067039109,
+ "loss": 1.988,
+ "step": 104
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017067039106145253,
+ "loss": 2.1008,
+ "step": 105
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017039106145251396,
+ "loss": 2.1482,
+ "step": 106
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00017011173184357543,
+ "loss": 2.1052,
+ "step": 107
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00016983240223463687,
+ "loss": 2.0978,
+ "step": 108
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016955307262569834,
+ "loss": 2.1303,
+ "step": 109
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016927374301675978,
+ "loss": 2.0794,
+ "step": 110
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.00016899441340782122,
+ "loss": 2.1059,
+ "step": 111
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0001687150837988827,
+ "loss": 1.9642,
+ "step": 112
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.00016843575418994415,
+ "loss": 2.0415,
+ "step": 113
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0001681564245810056,
+ "loss": 2.0795,
+ "step": 114
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.00016787709497206705,
+ "loss": 2.0238,
+ "step": 115
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0001675977653631285,
+ "loss": 2.078,
+ "step": 116
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.00016731843575418996,
+ "loss": 2.1362,
+ "step": 117
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0001670391061452514,
+ "loss": 2.0552,
+ "step": 118
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.00016675977653631284,
+ "loss": 2.213,
+ "step": 119
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0001664804469273743,
+ "loss": 1.9999,
+ "step": 120
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.00016620111731843577,
+ "loss": 2.06,
+ "step": 121
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0001659217877094972,
+ "loss": 2.0177,
+ "step": 122
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016564245810055868,
+ "loss": 2.0504,
+ "step": 123
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016536312849162012,
+ "loss": 2.0585,
+ "step": 124
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016508379888268158,
+ "loss": 2.0273,
+ "step": 125
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016480446927374302,
+ "loss": 2.0549,
+ "step": 126
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.00016452513966480446,
+ "loss": 2.044,
+ "step": 127
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016424581005586593,
+ "loss": 2.0731,
+ "step": 128
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016396648044692737,
+ "loss": 2.0568,
+ "step": 129
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.00016368715083798883,
+ "loss": 2.007,
+ "step": 130
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0001634078212290503,
+ "loss": 2.0707,
+ "step": 131
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.00016312849162011174,
+ "loss": 1.9793,
+ "step": 132
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0001628491620111732,
+ "loss": 2.1311,
+ "step": 133
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016256983240223465,
+ "loss": 2.0016,
+ "step": 134
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016229050279329609,
+ "loss": 1.9945,
+ "step": 135
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.00016201117318435755,
+ "loss": 2.0186,
+ "step": 136
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.000161731843575419,
+ "loss": 2.0971,
+ "step": 137
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.00016145251396648046,
+ "loss": 2.0883,
+ "step": 138
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016117318435754192,
+ "loss": 2.0803,
+ "step": 139
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016089385474860336,
+ "loss": 2.0617,
+ "step": 140
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016061452513966483,
+ "loss": 2.1265,
+ "step": 141
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016033519553072627,
+ "loss": 2.0151,
+ "step": 142
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0001600558659217877,
+ "loss": 1.996,
+ "step": 143
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.00015977653631284918,
+ "loss": 2.0164,
+ "step": 144
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.00015949720670391061,
+ "loss": 2.0314,
+ "step": 145
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015921787709497208,
+ "loss": 1.9501,
+ "step": 146
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015893854748603352,
+ "loss": 2.087,
+ "step": 147
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.000158659217877095,
+ "loss": 2.0262,
+ "step": 148
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.00015837988826815643,
+ "loss": 2.0765,
+ "step": 149
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0001581005586592179,
+ "loss": 2.105,
+ "step": 150
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.00015782122905027933,
+ "loss": 1.9863,
+ "step": 151
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0001575418994413408,
+ "loss": 1.9873,
+ "step": 152
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.00015726256983240224,
+ "loss": 2.0094,
+ "step": 153
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.00015698324022346368,
+ "loss": 1.9141,
+ "step": 154
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015670391061452514,
+ "loss": 1.917,
+ "step": 155
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015642458100558658,
+ "loss": 2.109,
+ "step": 156
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015614525139664805,
+ "loss": 1.9799,
+ "step": 157
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015586592178770952,
+ "loss": 1.9571,
+ "step": 158
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015558659217877096,
+ "loss": 1.9931,
+ "step": 159
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015530726256983242,
+ "loss": 2.1004,
+ "step": 160
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.00015502793296089386,
+ "loss": 2.0385,
+ "step": 161
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0001547486033519553,
+ "loss": 1.9751,
+ "step": 162
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.00015446927374301677,
+ "loss": 2.0544,
+ "step": 163
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0001541899441340782,
+ "loss": 2.0069,
+ "step": 164
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.00015391061452513967,
+ "loss": 1.9576,
+ "step": 165
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015363128491620114,
+ "loss": 1.8991,
+ "step": 166
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015335195530726258,
+ "loss": 1.9336,
+ "step": 167
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015307262569832405,
+ "loss": 1.9736,
+ "step": 168
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015279329608938548,
+ "loss": 1.9702,
+ "step": 169
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.00015251396648044692,
+ "loss": 1.9055,
+ "step": 170
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0001522346368715084,
+ "loss": 2.0503,
+ "step": 171
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.00015195530726256983,
+ "loss": 2.0039,
+ "step": 172
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0001516759776536313,
+ "loss": 1.9406,
+ "step": 173
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.00015139664804469274,
+ "loss": 2.0525,
+ "step": 174
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0001511173184357542,
+ "loss": 1.9234,
+ "step": 175
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.00015083798882681567,
+ "loss": 1.8614,
+ "step": 176
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0001505586592178771,
+ "loss": 1.9616,
+ "step": 177
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.00015027932960893855,
+ "loss": 1.9509,
+ "step": 178
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.9592,
+ "step": 179
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014972067039106145,
+ "loss": 1.8991,
+ "step": 180
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014944134078212292,
+ "loss": 1.9127,
+ "step": 181
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.00014916201117318436,
+ "loss": 1.8982,
+ "step": 182
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0001488826815642458,
+ "loss": 1.9534,
+ "step": 183
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0001486033519553073,
+ "loss": 1.7794,
+ "step": 184
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.00014832402234636873,
+ "loss": 1.7958,
+ "step": 185
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014804469273743017,
+ "loss": 1.8282,
+ "step": 186
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014776536312849164,
+ "loss": 2.0423,
+ "step": 187
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.00014748603351955308,
+ "loss": 1.9282,
+ "step": 188
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014720670391061454,
+ "loss": 1.9072,
+ "step": 189
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014692737430167598,
+ "loss": 1.8665,
+ "step": 190
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.00014664804469273742,
+ "loss": 1.9021,
+ "step": 191
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.0001463687150837989,
+ "loss": 1.7308,
+ "step": 192
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.00014608938547486035,
+ "loss": 1.9165,
+ "step": 193
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.0001458100558659218,
+ "loss": 1.842,
+ "step": 194
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.00014553072625698326,
+ "loss": 1.9128,
+ "step": 195
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.0001452513966480447,
+ "loss": 1.8005,
+ "step": 196
+ },
+ {
+ "epoch": 1.1,
+ "learning_rate": 0.00014497206703910614,
+ "loss": 1.8547,
+ "step": 197
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.0001446927374301676,
+ "loss": 1.9042,
+ "step": 198
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.00014441340782122905,
+ "loss": 1.8609,
+ "step": 199
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.0001441340782122905,
+ "loss": 1.9591,
+ "step": 200
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.00014385474860335195,
+ "loss": 1.8722,
+ "step": 201
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014357541899441342,
+ "loss": 1.8535,
+ "step": 202
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014329608938547488,
+ "loss": 1.8676,
+ "step": 203
+ },
+ {
+ "epoch": 1.14,
+ "learning_rate": 0.00014301675977653632,
+ "loss": 1.8976,
+ "step": 204
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014273743016759776,
+ "loss": 1.7723,
+ "step": 205
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014245810055865923,
+ "loss": 1.8165,
+ "step": 206
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014217877094972067,
+ "loss": 1.7811,
+ "step": 207
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014189944134078214,
+ "loss": 1.908,
+ "step": 208
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014162011173184357,
+ "loss": 1.7663,
+ "step": 209
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014134078212290501,
+ "loss": 1.6779,
+ "step": 210
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.0001410614525139665,
+ "loss": 1.9039,
+ "step": 211
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.00014078212290502795,
+ "loss": 1.8033,
+ "step": 212
+ },
+ {
+ "epoch": 1.19,
+ "learning_rate": 0.00014050279329608939,
+ "loss": 1.8251,
+ "step": 213
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.00014022346368715085,
+ "loss": 1.8505,
+ "step": 214
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.0001399441340782123,
+ "loss": 1.8147,
+ "step": 215
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.00013966480446927376,
+ "loss": 1.8419,
+ "step": 216
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.0001393854748603352,
+ "loss": 1.8401,
+ "step": 217
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.00013910614525139664,
+ "loss": 1.8912,
+ "step": 218
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.0001388268156424581,
+ "loss": 1.7548,
+ "step": 219
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.00013854748603351957,
+ "loss": 1.8741,
+ "step": 220
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.000138268156424581,
+ "loss": 1.9549,
+ "step": 221
+ },
+ {
+ "epoch": 1.24,
+ "learning_rate": 0.00013798882681564248,
+ "loss": 1.9093,
+ "step": 222
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013770949720670392,
+ "loss": 1.7896,
+ "step": 223
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013743016759776538,
+ "loss": 1.8491,
+ "step": 224
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013715083798882682,
+ "loss": 1.7851,
+ "step": 225
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013687150837988826,
+ "loss": 1.6992,
+ "step": 226
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013659217877094973,
+ "loss": 1.9765,
+ "step": 227
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013631284916201117,
+ "loss": 1.8179,
+ "step": 228
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.00013603351955307263,
+ "loss": 1.8548,
+ "step": 229
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.0001357541899441341,
+ "loss": 1.8843,
+ "step": 230
+ },
+ {
+ "epoch": 1.29,
+ "learning_rate": 0.00013547486033519554,
+ "loss": 1.9105,
+ "step": 231
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013519553072625698,
+ "loss": 1.8748,
+ "step": 232
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013491620111731844,
+ "loss": 1.7976,
+ "step": 233
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013463687150837988,
+ "loss": 1.7369,
+ "step": 234
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013435754189944135,
+ "loss": 1.7808,
+ "step": 235
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.0001340782122905028,
+ "loss": 1.8385,
+ "step": 236
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.00013379888268156423,
+ "loss": 1.8295,
+ "step": 237
+ },
+ {
+ "epoch": 1.33,
+ "learning_rate": 0.00013351955307262572,
+ "loss": 1.757,
+ "step": 238
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.00013324022346368716,
+ "loss": 1.7904,
+ "step": 239
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.0001329608938547486,
+ "loss": 1.7632,
+ "step": 240
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.00013268156424581007,
+ "loss": 1.7867,
+ "step": 241
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.0001324022346368715,
+ "loss": 1.8259,
+ "step": 242
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.00013212290502793297,
+ "loss": 1.6655,
+ "step": 243
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.0001318435754189944,
+ "loss": 1.6848,
+ "step": 244
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013156424581005585,
+ "loss": 1.7931,
+ "step": 245
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013128491620111732,
+ "loss": 1.7868,
+ "step": 246
+ },
+ {
+ "epoch": 1.38,
+ "learning_rate": 0.00013100558659217879,
+ "loss": 1.7732,
+ "step": 247
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.00013072625698324022,
+ "loss": 1.7851,
+ "step": 248
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.0001304469273743017,
+ "loss": 1.7406,
+ "step": 249
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.00013016759776536313,
+ "loss": 1.5853,
+ "step": 250
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.0001298882681564246,
+ "loss": 1.8271,
+ "step": 251
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012960893854748604,
+ "loss": 1.6054,
+ "step": 252
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012932960893854748,
+ "loss": 1.6884,
+ "step": 253
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012905027932960894,
+ "loss": 1.7333,
+ "step": 254
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012877094972067038,
+ "loss": 1.803,
+ "step": 255
+ },
+ {
+ "epoch": 1.43,
+ "learning_rate": 0.00012849162011173185,
+ "loss": 1.63,
+ "step": 256
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012821229050279331,
+ "loss": 1.8023,
+ "step": 257
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012793296089385475,
+ "loss": 1.762,
+ "step": 258
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012765363128491622,
+ "loss": 1.7565,
+ "step": 259
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012737430167597766,
+ "loss": 1.5937,
+ "step": 260
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.0001270949720670391,
+ "loss": 1.7208,
+ "step": 261
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.00012681564245810057,
+ "loss": 1.8097,
+ "step": 262
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.000126536312849162,
+ "loss": 1.7042,
+ "step": 263
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.00012625698324022347,
+ "loss": 1.7892,
+ "step": 264
+ },
+ {
+ "epoch": 1.48,
+ "learning_rate": 0.00012597765363128494,
+ "loss": 1.752,
+ "step": 265
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012569832402234638,
+ "loss": 1.7333,
+ "step": 266
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012541899441340784,
+ "loss": 1.7492,
+ "step": 267
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012513966480446928,
+ "loss": 1.818,
+ "step": 268
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012486033519553072,
+ "loss": 1.7146,
+ "step": 269
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.0001245810055865922,
+ "loss": 1.7958,
+ "step": 270
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.00012430167597765363,
+ "loss": 1.7212,
+ "step": 271
+ },
+ {
+ "epoch": 1.52,
+ "learning_rate": 0.0001240223463687151,
+ "loss": 1.8835,
+ "step": 272
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.00012374301675977656,
+ "loss": 1.7486,
+ "step": 273
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.000123463687150838,
+ "loss": 1.8103,
+ "step": 274
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.00012318435754189944,
+ "loss": 1.7557,
+ "step": 275
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.0001229050279329609,
+ "loss": 1.6198,
+ "step": 276
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.00012262569832402235,
+ "loss": 1.6971,
+ "step": 277
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.0001223463687150838,
+ "loss": 1.668,
+ "step": 278
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.00012206703910614525,
+ "loss": 1.8795,
+ "step": 279
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.0001217877094972067,
+ "loss": 1.6412,
+ "step": 280
+ },
+ {
+ "epoch": 1.57,
+ "learning_rate": 0.00012150837988826816,
+ "loss": 1.7497,
+ "step": 281
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012122905027932962,
+ "loss": 1.5577,
+ "step": 282
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012094972067039108,
+ "loss": 1.8049,
+ "step": 283
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012067039106145253,
+ "loss": 1.6834,
+ "step": 284
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012039106145251397,
+ "loss": 1.7978,
+ "step": 285
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00012011173184357542,
+ "loss": 1.6558,
+ "step": 286
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00011983240223463687,
+ "loss": 1.7561,
+ "step": 287
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011955307262569833,
+ "loss": 1.6807,
+ "step": 288
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011927374301675978,
+ "loss": 1.6937,
+ "step": 289
+ },
+ {
+ "epoch": 1.62,
+ "learning_rate": 0.00011899441340782122,
+ "loss": 1.6591,
+ "step": 290
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.0001187150837988827,
+ "loss": 1.6771,
+ "step": 291
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.00011843575418994415,
+ "loss": 1.7743,
+ "step": 292
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011815642458100559,
+ "loss": 1.5857,
+ "step": 293
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011787709497206705,
+ "loss": 1.6999,
+ "step": 294
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.0001175977653631285,
+ "loss": 1.5661,
+ "step": 295
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.00011731843575418995,
+ "loss": 1.7235,
+ "step": 296
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.0001170391061452514,
+ "loss": 1.607,
+ "step": 297
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.00011675977653631284,
+ "loss": 1.68,
+ "step": 298
+ },
+ {
+ "epoch": 1.67,
+ "learning_rate": 0.0001164804469273743,
+ "loss": 1.6938,
+ "step": 299
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011620111731843578,
+ "loss": 1.6315,
+ "step": 300
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011592178770949722,
+ "loss": 1.6802,
+ "step": 301
+ },
+ {
+ "epoch": 1.69,
+ "learning_rate": 0.00011564245810055867,
+ "loss": 1.7174,
+ "step": 302
+ },
+ {
+ "epoch": 1.69,
+ "learning_rate": 0.00011536312849162012,
+ "loss": 1.5212,
+ "step": 303
+ },
+ {
+ "epoch": 1.7,
+ "learning_rate": 0.00011508379888268157,
+ "loss": 1.5808,
+ "step": 304
+ },
+ {
+ "epoch": 1.7,
+ "learning_rate": 0.00011480446927374303,
+ "loss": 1.6152,
+ "step": 305
+ },
+ {
+ "epoch": 1.71,
+ "learning_rate": 0.00011452513966480447,
+ "loss": 1.5435,
+ "step": 306
+ },
+ {
+ "epoch": 1.72,
+ "learning_rate": 0.00011424581005586592,
+ "loss": 1.6603,
+ "step": 307
+ },
+ {
+ "epoch": 1.72,
+ "learning_rate": 0.00011396648044692737,
+ "loss": 1.685,
+ "step": 308
+ },
+ {
+ "epoch": 1.73,
+ "learning_rate": 0.00011368715083798884,
+ "loss": 1.6002,
+ "step": 309
+ },
+ {
+ "epoch": 1.73,
+ "learning_rate": 0.00011340782122905029,
+ "loss": 1.6046,
+ "step": 310
+ },
+ {
+ "epoch": 1.74,
+ "learning_rate": 0.00011312849162011174,
+ "loss": 1.5969,
+ "step": 311
+ },
+ {
+ "epoch": 1.74,
+ "learning_rate": 0.0001128491620111732,
+ "loss": 1.5845,
+ "step": 312
+ },
+ {
+ "epoch": 1.75,
+ "learning_rate": 0.00011256983240223464,
+ "loss": 1.8183,
+ "step": 313
+ },
+ {
+ "epoch": 1.75,
+ "learning_rate": 0.00011229050279329609,
+ "loss": 1.6953,
+ "step": 314
+ },
+ {
+ "epoch": 1.76,
+ "learning_rate": 0.00011201117318435754,
+ "loss": 1.7787,
+ "step": 315
+ },
+ {
+ "epoch": 1.77,
+ "learning_rate": 0.000111731843575419,
+ "loss": 1.6422,
+ "step": 316
+ },
+ {
+ "epoch": 1.77,
+ "learning_rate": 0.00011145251396648045,
+ "loss": 1.7034,
+ "step": 317
+ },
+ {
+ "epoch": 1.78,
+ "learning_rate": 0.00011117318435754192,
+ "loss": 1.7301,
+ "step": 318
+ },
+ {
+ "epoch": 1.78,
+ "learning_rate": 0.00011089385474860337,
+ "loss": 1.7084,
+ "step": 319
+ },
+ {
+ "epoch": 1.79,
+ "learning_rate": 0.00011061452513966482,
+ "loss": 1.772,
+ "step": 320
+ },
+ {
+ "epoch": 1.79,
+ "learning_rate": 0.00011033519553072626,
+ "loss": 1.5733,
+ "step": 321
+ },
+ {
+ "epoch": 1.8,
+ "learning_rate": 0.00011005586592178771,
+ "loss": 1.6423,
+ "step": 322
+ },
+ {
+ "epoch": 1.8,
+ "learning_rate": 0.00010977653631284917,
+ "loss": 1.5809,
+ "step": 323
+ },
+ {
+ "epoch": 1.81,
+ "learning_rate": 0.00010949720670391062,
+ "loss": 1.6781,
+ "step": 324
+ },
+ {
+ "epoch": 1.82,
+ "learning_rate": 0.00010921787709497207,
+ "loss": 1.6788,
+ "step": 325
+ },
+ {
+ "epoch": 1.82,
+ "learning_rate": 0.00010893854748603351,
+ "loss": 1.6346,
+ "step": 326
+ },
+ {
+ "epoch": 1.83,
+ "learning_rate": 0.00010865921787709499,
+ "loss": 1.6634,
+ "step": 327
+ },
+ {
+ "epoch": 1.83,
+ "learning_rate": 0.00010837988826815643,
+ "loss": 1.7561,
+ "step": 328
+ },
+ {
+ "epoch": 1.84,
+ "learning_rate": 0.00010810055865921788,
+ "loss": 1.66,
+ "step": 329
+ },
+ {
+ "epoch": 1.84,
+ "learning_rate": 0.00010782122905027934,
+ "loss": 1.7298,
+ "step": 330
+ },
+ {
+ "epoch": 1.85,
+ "learning_rate": 0.00010754189944134079,
+ "loss": 1.6893,
+ "step": 331
+ },
+ {
+ "epoch": 1.85,
+ "learning_rate": 0.00010726256983240224,
+ "loss": 1.7631,
+ "step": 332
+ },
+ {
+ "epoch": 1.86,
+ "learning_rate": 0.00010698324022346368,
+ "loss": 1.6633,
+ "step": 333
+ },
+ {
+ "epoch": 1.87,
+ "learning_rate": 0.00010670391061452513,
+ "loss": 1.5388,
+ "step": 334
+ },
+ {
+ "epoch": 1.87,
+ "learning_rate": 0.00010642458100558659,
+ "loss": 1.6718,
+ "step": 335
+ },
+ {
+ "epoch": 1.88,
+ "learning_rate": 0.00010614525139664805,
+ "loss": 1.5536,
+ "step": 336
+ },
+ {
+ "epoch": 1.88,
+ "learning_rate": 0.00010586592178770951,
+ "loss": 1.6483,
+ "step": 337
+ },
+ {
+ "epoch": 1.89,
+ "learning_rate": 0.00010558659217877096,
+ "loss": 1.5774,
+ "step": 338
+ },
+ {
+ "epoch": 1.89,
+ "learning_rate": 0.00010530726256983241,
+ "loss": 1.6366,
+ "step": 339
+ },
+ {
+ "epoch": 1.9,
+ "learning_rate": 0.00010502793296089387,
+ "loss": 1.5567,
+ "step": 340
+ },
+ {
+ "epoch": 1.91,
+ "learning_rate": 0.0001047486033519553,
+ "loss": 1.5323,
+ "step": 341
+ },
+ {
+ "epoch": 1.91,
+ "learning_rate": 0.00010446927374301676,
+ "loss": 1.4608,
+ "step": 342
+ },
+ {
+ "epoch": 1.92,
+ "learning_rate": 0.00010418994413407821,
+ "loss": 1.5933,
+ "step": 343
+ },
+ {
+ "epoch": 1.92,
+ "learning_rate": 0.00010391061452513966,
+ "loss": 1.6625,
+ "step": 344
+ },
+ {
+ "epoch": 1.93,
+ "learning_rate": 0.00010363128491620113,
+ "loss": 1.7236,
+ "step": 345
+ },
+ {
+ "epoch": 1.93,
+ "learning_rate": 0.00010335195530726258,
+ "loss": 1.759,
+ "step": 346
+ },
+ {
+ "epoch": 1.94,
+ "learning_rate": 0.00010307262569832404,
+ "loss": 1.7248,
+ "step": 347
+ },
+ {
+ "epoch": 1.94,
+ "learning_rate": 0.00010279329608938548,
+ "loss": 1.5144,
+ "step": 348
+ },
+ {
+ "epoch": 1.95,
+ "learning_rate": 0.00010251396648044693,
+ "loss": 1.6905,
+ "step": 349
+ },
+ {
+ "epoch": 1.96,
+ "learning_rate": 0.00010223463687150838,
+ "loss": 1.6119,
+ "step": 350
+ },
+ {
+ "epoch": 1.96,
+ "learning_rate": 0.00010195530726256983,
+ "loss": 1.5464,
+ "step": 351
+ },
+ {
+ "epoch": 1.97,
+ "learning_rate": 0.00010167597765363129,
+ "loss": 1.6901,
+ "step": 352
+ },
+ {
+ "epoch": 1.97,
+ "learning_rate": 0.00010139664804469273,
+ "loss": 1.3511,
+ "step": 353
+ },
+ {
+ "epoch": 1.98,
+ "learning_rate": 0.0001011173184357542,
+ "loss": 1.5434,
+ "step": 354
+ },
+ {
+ "epoch": 1.98,
+ "learning_rate": 0.00010083798882681566,
+ "loss": 1.5891,
+ "step": 355
+ },
+ {
+ "epoch": 1.99,
+ "learning_rate": 0.0001005586592178771,
+ "loss": 1.6658,
+ "step": 356
+ },
+ {
+ "epoch": 1.99,
+ "learning_rate": 0.00010027932960893855,
+ "loss": 1.5657,
+ "step": 357
+ },
+ {
+ "epoch": 2.0,
+ "learning_rate": 0.0001,
+ "loss": 1.7005,
+ "step": 358
+ },
+ {
+ "epoch": 2.01,
+ "learning_rate": 9.972067039106146e-05,
+ "loss": 1.4202,
+ "step": 359
+ },
+ {
+ "epoch": 2.01,
+ "learning_rate": 9.944134078212291e-05,
+ "loss": 1.5262,
+ "step": 360
+ },
+ {
+ "epoch": 2.02,
+ "learning_rate": 9.916201117318436e-05,
+ "loss": 1.6323,
+ "step": 361
+ },
+ {
+ "epoch": 2.02,
+ "learning_rate": 9.888268156424582e-05,
+ "loss": 1.5521,
+ "step": 362
+ },
+ {
+ "epoch": 2.03,
+ "learning_rate": 9.860335195530727e-05,
+ "loss": 1.5762,
+ "step": 363
+ },
+ {
+ "epoch": 2.03,
+ "learning_rate": 9.832402234636872e-05,
+ "loss": 1.613,
+ "step": 364
+ },
+ {
+ "epoch": 2.04,
+ "learning_rate": 9.804469273743018e-05,
+ "loss": 1.4231,
+ "step": 365
+ },
+ {
+ "epoch": 2.04,
+ "learning_rate": 9.776536312849163e-05,
+ "loss": 1.5706,
+ "step": 366
+ },
+ {
+ "epoch": 2.05,
+ "learning_rate": 9.748603351955308e-05,
+ "loss": 1.5245,
+ "step": 367
+ },
+ {
+ "epoch": 2.06,
+ "learning_rate": 9.720670391061453e-05,
+ "loss": 1.4771,
+ "step": 368
+ },
+ {
+ "epoch": 2.06,
+ "learning_rate": 9.692737430167597e-05,
+ "loss": 1.596,
+ "step": 369
+ },
+ {
+ "epoch": 2.07,
+ "learning_rate": 9.664804469273744e-05,
+ "loss": 1.537,
+ "step": 370
+ },
+ {
+ "epoch": 2.07,
+ "learning_rate": 9.636871508379889e-05,
+ "loss": 1.4276,
+ "step": 371
+ },
+ {
+ "epoch": 2.08,
+ "learning_rate": 9.608938547486033e-05,
+ "loss": 1.4746,
+ "step": 372
+ },
+ {
+ "epoch": 2.08,
+ "learning_rate": 9.581005586592178e-05,
+ "loss": 1.4374,
+ "step": 373
+ },
+ {
+ "epoch": 2.09,
+ "learning_rate": 9.553072625698325e-05,
+ "loss": 1.4704,
+ "step": 374
+ },
+ {
+ "epoch": 2.09,
+ "learning_rate": 9.52513966480447e-05,
+ "loss": 1.5997,
+ "step": 375
+ },
+ {
+ "epoch": 2.1,
+ "learning_rate": 9.497206703910614e-05,
+ "loss": 1.5034,
+ "step": 376
+ },
+ {
+ "epoch": 2.11,
+ "learning_rate": 9.46927374301676e-05,
+ "loss": 1.6392,
+ "step": 377
+ },
+ {
+ "epoch": 2.11,
+ "learning_rate": 9.441340782122905e-05,
+ "loss": 1.5611,
+ "step": 378
+ },
+ {
+ "epoch": 2.12,
+ "learning_rate": 9.413407821229052e-05,
+ "loss": 1.3384,
+ "step": 379
+ },
+ {
+ "epoch": 2.12,
+ "learning_rate": 9.385474860335196e-05,
+ "loss": 1.5732,
+ "step": 380
+ },
+ {
+ "epoch": 2.13,
+ "learning_rate": 9.357541899441341e-05,
+ "loss": 1.3874,
+ "step": 381
+ },
+ {
+ "epoch": 2.13,
+ "learning_rate": 9.329608938547486e-05,
+ "loss": 1.4139,
+ "step": 382
+ },
+ {
+ "epoch": 2.14,
+ "learning_rate": 9.301675977653633e-05,
+ "loss": 1.5195,
+ "step": 383
+ },
+ {
+ "epoch": 2.15,
+ "learning_rate": 9.273743016759777e-05,
+ "loss": 1.4371,
+ "step": 384
+ },
+ {
+ "epoch": 2.15,
+ "learning_rate": 9.245810055865922e-05,
+ "loss": 1.4411,
+ "step": 385
+ },
+ {
+ "epoch": 2.16,
+ "learning_rate": 9.217877094972067e-05,
+ "loss": 1.5214,
+ "step": 386
+ },
+ {
+ "epoch": 2.16,
+ "learning_rate": 9.189944134078213e-05,
+ "loss": 1.4971,
+ "step": 387
+ },
+ {
+ "epoch": 2.17,
+ "learning_rate": 9.162011173184358e-05,
+ "loss": 1.3453,
+ "step": 388
+ },
+ {
+ "epoch": 2.17,
+ "learning_rate": 9.134078212290503e-05,
+ "loss": 1.394,
+ "step": 389
+ },
+ {
+ "epoch": 2.18,
+ "learning_rate": 9.106145251396648e-05,
+ "loss": 1.5058,
+ "step": 390
+ },
+ {
+ "epoch": 2.18,
+ "learning_rate": 9.078212290502794e-05,
+ "loss": 1.4855,
+ "step": 391
+ },
+ {
+ "epoch": 2.19,
+ "learning_rate": 9.050279329608939e-05,
+ "loss": 1.4647,
+ "step": 392
+ },
+ {
+ "epoch": 2.2,
+ "learning_rate": 9.022346368715084e-05,
+ "loss": 1.3435,
+ "step": 393
+ },
+ {
+ "epoch": 2.2,
+ "learning_rate": 8.99441340782123e-05,
+ "loss": 1.5815,
+ "step": 394
+ },
+ {
+ "epoch": 2.21,
+ "learning_rate": 8.966480446927375e-05,
+ "loss": 1.4742,
+ "step": 395
+ },
+ {
+ "epoch": 2.21,
+ "learning_rate": 8.938547486033519e-05,
+ "loss": 1.5389,
+ "step": 396
+ },
+ {
+ "epoch": 2.22,
+ "learning_rate": 8.910614525139666e-05,
+ "loss": 1.384,
+ "step": 397
+ },
+ {
+ "epoch": 2.22,
+ "learning_rate": 8.882681564245811e-05,
+ "loss": 1.3967,
+ "step": 398
+ },
+ {
+ "epoch": 2.23,
+ "learning_rate": 8.854748603351956e-05,
+ "loss": 1.442,
+ "step": 399
+ },
+ {
+ "epoch": 2.23,
+ "learning_rate": 8.8268156424581e-05,
+ "loss": 1.396,
+ "step": 400
+ },
+ {
+ "epoch": 2.24,
+ "learning_rate": 8.798882681564247e-05,
+ "loss": 1.4706,
+ "step": 401
+ },
+ {
+ "epoch": 2.25,
+ "learning_rate": 8.770949720670392e-05,
+ "loss": 1.4217,
+ "step": 402
+ },
+ {
+ "epoch": 2.25,
+ "learning_rate": 8.743016759776537e-05,
+ "loss": 1.5617,
+ "step": 403
+ },
+ {
+ "epoch": 2.26,
+ "learning_rate": 8.715083798882681e-05,
+ "loss": 1.5701,
+ "step": 404
+ },
+ {
+ "epoch": 2.26,
+ "learning_rate": 8.687150837988828e-05,
+ "loss": 1.3659,
+ "step": 405
+ },
+ {
+ "epoch": 2.27,
+ "learning_rate": 8.659217877094973e-05,
+ "loss": 1.3424,
+ "step": 406
+ },
+ {
+ "epoch": 2.27,
+ "learning_rate": 8.631284916201118e-05,
+ "loss": 1.5782,
+ "step": 407
+ },
+ {
+ "epoch": 2.28,
+ "learning_rate": 8.603351955307262e-05,
+ "loss": 1.3355,
+ "step": 408
+ },
+ {
+ "epoch": 2.28,
+ "learning_rate": 8.575418994413408e-05,
+ "loss": 1.4798,
+ "step": 409
+ },
+ {
+ "epoch": 2.29,
+ "learning_rate": 8.547486033519554e-05,
+ "loss": 1.3813,
+ "step": 410
+ },
+ {
+ "epoch": 2.3,
+ "learning_rate": 8.519553072625698e-05,
+ "loss": 1.6011,
+ "step": 411
+ },
+ {
+ "epoch": 2.3,
+ "learning_rate": 8.491620111731844e-05,
+ "loss": 1.4596,
+ "step": 412
+ },
+ {
+ "epoch": 2.31,
+ "learning_rate": 8.463687150837989e-05,
+ "loss": 1.4997,
+ "step": 413
+ },
+ {
+ "epoch": 2.31,
+ "learning_rate": 8.435754189944135e-05,
+ "loss": 1.5255,
+ "step": 414
+ },
+ {
+ "epoch": 2.32,
+ "learning_rate": 8.40782122905028e-05,
+ "loss": 1.5305,
+ "step": 415
+ },
+ {
+ "epoch": 2.32,
+ "learning_rate": 8.379888268156425e-05,
+ "loss": 1.3029,
+ "step": 416
+ },
+ {
+ "epoch": 2.33,
+ "learning_rate": 8.35195530726257e-05,
+ "loss": 1.2046,
+ "step": 417
+ },
+ {
+ "epoch": 2.34,
+ "learning_rate": 8.324022346368715e-05,
+ "loss": 1.4985,
+ "step": 418
+ },
+ {
+ "epoch": 2.34,
+ "learning_rate": 8.29608938547486e-05,
+ "loss": 1.3486,
+ "step": 419
+ },
+ {
+ "epoch": 2.35,
+ "learning_rate": 8.268156424581006e-05,
+ "loss": 1.2839,
+ "step": 420
+ },
+ {
+ "epoch": 2.35,
+ "learning_rate": 8.240223463687151e-05,
+ "loss": 1.4286,
+ "step": 421
+ },
+ {
+ "epoch": 2.36,
+ "learning_rate": 8.212290502793296e-05,
+ "loss": 1.3893,
+ "step": 422
+ },
+ {
+ "epoch": 2.36,
+ "learning_rate": 8.184357541899442e-05,
+ "loss": 1.5477,
+ "step": 423
+ },
+ {
+ "epoch": 2.37,
+ "learning_rate": 8.156424581005587e-05,
+ "loss": 1.4973,
+ "step": 424
+ },
+ {
+ "epoch": 2.37,
+ "learning_rate": 8.128491620111732e-05,
+ "loss": 1.3549,
+ "step": 425
+ },
+ {
+ "epoch": 2.38,
+ "learning_rate": 8.100558659217878e-05,
+ "loss": 1.3966,
+ "step": 426
+ },
+ {
+ "epoch": 2.39,
+ "learning_rate": 8.072625698324023e-05,
+ "loss": 1.3974,
+ "step": 427
+ },
+ {
+ "epoch": 2.39,
+ "learning_rate": 8.044692737430168e-05,
+ "loss": 1.5211,
+ "step": 428
+ },
+ {
+ "epoch": 2.4,
+ "learning_rate": 8.016759776536313e-05,
+ "loss": 1.4751,
+ "step": 429
+ },
+ {
+ "epoch": 2.4,
+ "learning_rate": 7.988826815642459e-05,
+ "loss": 1.4322,
+ "step": 430
+ },
+ {
+ "epoch": 2.41,
+ "learning_rate": 7.960893854748604e-05,
+ "loss": 1.5991,
+ "step": 431
+ },
+ {
+ "epoch": 2.41,
+ "learning_rate": 7.93296089385475e-05,
+ "loss": 1.3288,
+ "step": 432
+ },
+ {
+ "epoch": 2.42,
+ "learning_rate": 7.905027932960895e-05,
+ "loss": 1.4098,
+ "step": 433
+ },
+ {
+ "epoch": 2.42,
+ "learning_rate": 7.87709497206704e-05,
+ "loss": 1.4819,
+ "step": 434
+ },
+ {
+ "epoch": 2.43,
+ "learning_rate": 7.849162011173184e-05,
+ "loss": 1.4003,
+ "step": 435
+ },
+ {
+ "epoch": 2.44,
+ "learning_rate": 7.821229050279329e-05,
+ "loss": 1.3475,
+ "step": 436
+ },
+ {
+ "epoch": 2.44,
+ "learning_rate": 7.793296089385476e-05,
+ "loss": 1.3354,
+ "step": 437
+ },
+ {
+ "epoch": 2.45,
+ "learning_rate": 7.765363128491621e-05,
+ "loss": 1.3144,
+ "step": 438
+ },
+ {
+ "epoch": 2.45,
+ "learning_rate": 7.737430167597765e-05,
+ "loss": 1.4118,
+ "step": 439
+ },
+ {
+ "epoch": 2.46,
+ "learning_rate": 7.70949720670391e-05,
+ "loss": 1.3716,
+ "step": 440
+ },
+ {
+ "epoch": 2.46,
+ "learning_rate": 7.681564245810057e-05,
+ "loss": 1.4191,
+ "step": 441
+ },
+ {
+ "epoch": 2.47,
+ "learning_rate": 7.653631284916202e-05,
+ "loss": 1.5173,
+ "step": 442
+ },
+ {
+ "epoch": 2.47,
+ "learning_rate": 7.625698324022346e-05,
+ "loss": 1.3005,
+ "step": 443
+ },
+ {
+ "epoch": 2.48,
+ "learning_rate": 7.597765363128491e-05,
+ "loss": 1.4586,
+ "step": 444
+ },
+ {
+ "epoch": 2.49,
+ "learning_rate": 7.569832402234637e-05,
+ "loss": 1.5086,
+ "step": 445
+ },
+ {
+ "epoch": 2.49,
+ "learning_rate": 7.541899441340783e-05,
+ "loss": 1.5446,
+ "step": 446
+ },
+ {
+ "epoch": 2.5,
+ "learning_rate": 7.513966480446927e-05,
+ "loss": 1.3274,
+ "step": 447
+ },
+ {
+ "epoch": 2.5,
+ "learning_rate": 7.486033519553073e-05,
+ "loss": 1.4424,
+ "step": 448
+ },
+ {
+ "epoch": 2.51,
+ "learning_rate": 7.458100558659218e-05,
+ "loss": 1.2836,
+ "step": 449
+ },
+ {
+ "epoch": 2.51,
+ "learning_rate": 7.430167597765365e-05,
+ "loss": 1.4322,
+ "step": 450
+ },
+ {
+ "epoch": 2.52,
+ "learning_rate": 7.402234636871509e-05,
+ "loss": 1.4168,
+ "step": 451
+ },
+ {
+ "epoch": 2.53,
+ "learning_rate": 7.374301675977654e-05,
+ "loss": 1.3892,
+ "step": 452
+ },
+ {
+ "epoch": 2.53,
+ "learning_rate": 7.346368715083799e-05,
+ "loss": 1.6538,
+ "step": 453
+ },
+ {
+ "epoch": 2.54,
+ "learning_rate": 7.318435754189944e-05,
+ "loss": 1.4573,
+ "step": 454
+ },
+ {
+ "epoch": 2.54,
+ "learning_rate": 7.29050279329609e-05,
+ "loss": 1.4593,
+ "step": 455
+ },
+ {
+ "epoch": 2.55,
+ "learning_rate": 7.262569832402235e-05,
+ "loss": 1.4316,
+ "step": 456
+ },
+ {
+ "epoch": 2.55,
+ "learning_rate": 7.23463687150838e-05,
+ "loss": 1.2702,
+ "step": 457
+ },
+ {
+ "epoch": 2.56,
+ "learning_rate": 7.206703910614526e-05,
+ "loss": 1.5458,
+ "step": 458
+ },
+ {
+ "epoch": 2.56,
+ "learning_rate": 7.178770949720671e-05,
+ "loss": 1.5179,
+ "step": 459
+ },
+ {
+ "epoch": 2.57,
+ "learning_rate": 7.150837988826816e-05,
+ "loss": 1.3844,
+ "step": 460
+ },
+ {
+ "epoch": 2.58,
+ "learning_rate": 7.122905027932961e-05,
+ "loss": 1.3698,
+ "step": 461
+ },
+ {
+ "epoch": 2.58,
+ "learning_rate": 7.094972067039107e-05,
+ "loss": 1.4396,
+ "step": 462
+ },
+ {
+ "epoch": 2.59,
+ "learning_rate": 7.067039106145251e-05,
+ "loss": 1.5486,
+ "step": 463
+ },
+ {
+ "epoch": 2.59,
+ "learning_rate": 7.039106145251397e-05,
+ "loss": 1.2032,
+ "step": 464
+ },
+ {
+ "epoch": 2.6,
+ "learning_rate": 7.011173184357543e-05,
+ "loss": 1.3434,
+ "step": 465
+ },
+ {
+ "epoch": 2.6,
+ "learning_rate": 6.983240223463688e-05,
+ "loss": 1.4065,
+ "step": 466
+ },
+ {
+ "epoch": 2.61,
+ "learning_rate": 6.955307262569832e-05,
+ "loss": 1.3362,
+ "step": 467
+ },
+ {
+ "epoch": 2.61,
+ "learning_rate": 6.927374301675979e-05,
+ "loss": 1.375,
+ "step": 468
+ },
+ {
+ "epoch": 2.62,
+ "learning_rate": 6.899441340782124e-05,
+ "loss": 1.4549,
+ "step": 469
+ },
+ {
+ "epoch": 2.63,
+ "learning_rate": 6.871508379888269e-05,
+ "loss": 1.3724,
+ "step": 470
+ },
+ {
+ "epoch": 2.63,
+ "learning_rate": 6.843575418994413e-05,
+ "loss": 1.1602,
+ "step": 471
+ },
+ {
+ "epoch": 2.64,
+ "learning_rate": 6.815642458100558e-05,
+ "loss": 1.4346,
+ "step": 472
+ },
+ {
+ "epoch": 2.64,
+ "learning_rate": 6.787709497206705e-05,
+ "loss": 1.2877,
+ "step": 473
+ },
+ {
+ "epoch": 2.65,
+ "learning_rate": 6.759776536312849e-05,
+ "loss": 1.394,
+ "step": 474
+ },
+ {
+ "epoch": 2.65,
+ "learning_rate": 6.731843575418994e-05,
+ "loss": 1.3387,
+ "step": 475
+ },
+ {
+ "epoch": 2.66,
+ "learning_rate": 6.70391061452514e-05,
+ "loss": 1.41,
+ "step": 476
+ },
+ {
+ "epoch": 2.66,
+ "learning_rate": 6.675977653631286e-05,
+ "loss": 1.4299,
+ "step": 477
+ },
+ {
+ "epoch": 2.67,
+ "learning_rate": 6.64804469273743e-05,
+ "loss": 1.3824,
+ "step": 478
+ },
+ {
+ "epoch": 2.68,
+ "learning_rate": 6.620111731843575e-05,
+ "loss": 1.2656,
+ "step": 479
+ },
+ {
+ "epoch": 2.68,
+ "learning_rate": 6.59217877094972e-05,
+ "loss": 1.2811,
+ "step": 480
+ },
+ {
+ "epoch": 2.69,
+ "learning_rate": 6.564245810055866e-05,
+ "loss": 1.6839,
+ "step": 481
+ },
+ {
+ "epoch": 2.69,
+ "learning_rate": 6.536312849162011e-05,
+ "loss": 1.3089,
+ "step": 482
+ },
+ {
+ "epoch": 2.7,
+ "learning_rate": 6.508379888268157e-05,
+ "loss": 1.4323,
+ "step": 483
+ },
+ {
+ "epoch": 2.7,
+ "learning_rate": 6.480446927374302e-05,
+ "loss": 1.3439,
+ "step": 484
+ },
+ {
+ "epoch": 2.71,
+ "learning_rate": 6.452513966480447e-05,
+ "loss": 1.3614,
+ "step": 485
+ },
+ {
+ "epoch": 2.72,
+ "learning_rate": 6.424581005586592e-05,
+ "loss": 1.4974,
+ "step": 486
+ },
+ {
+ "epoch": 2.72,
+ "learning_rate": 6.396648044692738e-05,
+ "loss": 1.4963,
+ "step": 487
+ },
+ {
+ "epoch": 2.73,
+ "learning_rate": 6.368715083798883e-05,
+ "loss": 1.2902,
+ "step": 488
+ },
+ {
+ "epoch": 2.73,
+ "learning_rate": 6.340782122905028e-05,
+ "loss": 1.3226,
+ "step": 489
+ },
+ {
+ "epoch": 2.74,
+ "learning_rate": 6.312849162011174e-05,
+ "loss": 1.4353,
+ "step": 490
+ },
+ {
+ "epoch": 2.74,
+ "learning_rate": 6.284916201117319e-05,
+ "loss": 1.281,
+ "step": 491
+ },
+ {
+ "epoch": 2.75,
+ "learning_rate": 6.256983240223464e-05,
+ "loss": 1.3755,
+ "step": 492
+ },
+ {
+ "epoch": 2.75,
+ "learning_rate": 6.22905027932961e-05,
+ "loss": 1.3811,
+ "step": 493
+ },
+ {
+ "epoch": 2.76,
+ "learning_rate": 6.201117318435755e-05,
+ "loss": 1.4627,
+ "step": 494
+ },
+ {
+ "epoch": 2.77,
+ "learning_rate": 6.1731843575419e-05,
+ "loss": 1.3482,
+ "step": 495
+ },
+ {
+ "epoch": 2.77,
+ "learning_rate": 6.145251396648045e-05,
+ "loss": 1.4274,
+ "step": 496
+ },
+ {
+ "epoch": 2.78,
+ "learning_rate": 6.11731843575419e-05,
+ "loss": 1.3465,
+ "step": 497
+ },
+ {
+ "epoch": 2.78,
+ "learning_rate": 6.089385474860335e-05,
+ "loss": 1.2319,
+ "step": 498
+ },
+ {
+ "epoch": 2.79,
+ "learning_rate": 6.061452513966481e-05,
+ "loss": 1.2835,
+ "step": 499
+ },
+ {
+ "epoch": 2.79,
+ "learning_rate": 6.0335195530726265e-05,
+ "loss": 1.4836,
+ "step": 500
+ },
+ {
+ "epoch": 2.8,
+ "learning_rate": 6.005586592178771e-05,
+ "loss": 1.3878,
+ "step": 501
+ },
+ {
+ "epoch": 2.8,
+ "learning_rate": 5.9776536312849164e-05,
+ "loss": 1.4367,
+ "step": 502
+ },
+ {
+ "epoch": 2.81,
+ "learning_rate": 5.949720670391061e-05,
+ "loss": 1.2994,
+ "step": 503
+ },
+ {
+ "epoch": 2.82,
+ "learning_rate": 5.9217877094972076e-05,
+ "loss": 1.3599,
+ "step": 504
+ },
+ {
+ "epoch": 2.82,
+ "learning_rate": 5.893854748603352e-05,
+ "loss": 1.5357,
+ "step": 505
+ },
+ {
+ "epoch": 2.83,
+ "learning_rate": 5.8659217877094976e-05,
+ "loss": 1.4497,
+ "step": 506
+ },
+ {
+ "epoch": 2.83,
+ "learning_rate": 5.837988826815642e-05,
+ "loss": 1.3496,
+ "step": 507
+ },
+ {
+ "epoch": 2.84,
+ "learning_rate": 5.810055865921789e-05,
+ "loss": 1.294,
+ "step": 508
+ },
+ {
+ "epoch": 2.84,
+ "learning_rate": 5.7821229050279334e-05,
+ "loss": 1.4204,
+ "step": 509
+ },
+ {
+ "epoch": 2.85,
+ "learning_rate": 5.754189944134079e-05,
+ "loss": 1.5231,
+ "step": 510
+ },
+ {
+ "epoch": 2.85,
+ "learning_rate": 5.726256983240223e-05,
+ "loss": 1.3165,
+ "step": 511
+ },
+ {
+ "epoch": 2.86,
+ "learning_rate": 5.6983240223463686e-05,
+ "loss": 1.0796,
+ "step": 512
+ },
+ {
+ "epoch": 2.87,
+ "learning_rate": 5.6703910614525146e-05,
+ "loss": 1.4086,
+ "step": 513
+ },
+ {
+ "epoch": 2.87,
+ "learning_rate": 5.64245810055866e-05,
+ "loss": 1.2895,
+ "step": 514
+ },
+ {
+ "epoch": 2.88,
+ "learning_rate": 5.6145251396648045e-05,
+ "loss": 1.3024,
+ "step": 515
+ },
+ {
+ "epoch": 2.88,
+ "learning_rate": 5.58659217877095e-05,
+ "loss": 1.2479,
+ "step": 516
+ },
+ {
+ "epoch": 2.89,
+ "learning_rate": 5.558659217877096e-05,
+ "loss": 1.5232,
+ "step": 517
+ },
+ {
+ "epoch": 2.89,
+ "learning_rate": 5.530726256983241e-05,
+ "loss": 1.2654,
+ "step": 518
+ },
+ {
+ "epoch": 2.9,
+ "learning_rate": 5.502793296089386e-05,
+ "loss": 1.3018,
+ "step": 519
+ },
+ {
+ "epoch": 2.91,
+ "learning_rate": 5.474860335195531e-05,
+ "loss": 1.3077,
+ "step": 520
+ },
+ {
+ "epoch": 2.91,
+ "learning_rate": 5.4469273743016756e-05,
+ "loss": 1.3501,
+ "step": 521
+ },
+ {
+ "epoch": 2.92,
+ "learning_rate": 5.4189944134078215e-05,
+ "loss": 1.2607,
+ "step": 522
+ },
+ {
+ "epoch": 2.92,
+ "learning_rate": 5.391061452513967e-05,
+ "loss": 1.3819,
+ "step": 523
+ },
+ {
+ "epoch": 2.93,
+ "learning_rate": 5.363128491620112e-05,
+ "loss": 1.321,
+ "step": 524
+ },
+ {
+ "epoch": 2.93,
+ "learning_rate": 5.335195530726257e-05,
+ "loss": 1.3351,
+ "step": 525
+ },
+ {
+ "epoch": 2.94,
+ "learning_rate": 5.307262569832403e-05,
+ "loss": 1.2205,
+ "step": 526
+ },
+ {
+ "epoch": 2.94,
+ "learning_rate": 5.279329608938548e-05,
+ "loss": 1.2579,
+ "step": 527
+ },
+ {
+ "epoch": 2.95,
+ "learning_rate": 5.251396648044693e-05,
+ "loss": 1.2771,
+ "step": 528
+ },
+ {
+ "epoch": 2.96,
+ "learning_rate": 5.223463687150838e-05,
+ "loss": 1.3492,
+ "step": 529
+ },
+ {
+ "epoch": 2.96,
+ "learning_rate": 5.195530726256983e-05,
+ "loss": 1.3232,
+ "step": 530
+ },
+ {
+ "epoch": 2.97,
+ "learning_rate": 5.167597765363129e-05,
+ "loss": 1.2223,
+ "step": 531
+ },
+ {
+ "epoch": 2.97,
+ "learning_rate": 5.139664804469274e-05,
+ "loss": 1.3983,
+ "step": 532
+ },
+ {
+ "epoch": 2.98,
+ "learning_rate": 5.111731843575419e-05,
+ "loss": 1.2987,
+ "step": 533
+ },
+ {
+ "epoch": 2.98,
+ "learning_rate": 5.0837988826815644e-05,
+ "loss": 1.3717,
+ "step": 534
+ },
+ {
+ "epoch": 2.99,
+ "learning_rate": 5.05586592178771e-05,
+ "loss": 1.4467,
+ "step": 535
+ },
+ {
+ "epoch": 2.99,
+ "learning_rate": 5.027932960893855e-05,
+ "loss": 1.4314,
+ "step": 536
+ },
+ {
+ "epoch": 3.0,
+ "learning_rate": 5e-05,
+ "loss": 1.6593,
+ "step": 537
+ },
+ {
+ "epoch": 3.01,
+ "learning_rate": 4.9720670391061455e-05,
+ "loss": 1.3425,
+ "step": 538
+ },
+ {
+ "epoch": 3.01,
+ "learning_rate": 4.944134078212291e-05,
+ "loss": 1.4171,
+ "step": 539
+ },
+ {
+ "epoch": 3.02,
+ "learning_rate": 4.916201117318436e-05,
+ "loss": 1.3766,
+ "step": 540
+ },
+ {
+ "epoch": 3.02,
+ "learning_rate": 4.8882681564245814e-05,
+ "loss": 1.1605,
+ "step": 541
+ },
+ {
+ "epoch": 3.03,
+ "learning_rate": 4.860335195530727e-05,
+ "loss": 1.3132,
+ "step": 542
+ },
+ {
+ "epoch": 3.03,
+ "learning_rate": 4.832402234636872e-05,
+ "loss": 1.4128,
+ "step": 543
+ },
+ {
+ "epoch": 3.04,
+ "learning_rate": 4.8044692737430166e-05,
+ "loss": 1.4902,
+ "step": 544
+ },
+ {
+ "epoch": 3.04,
+ "learning_rate": 4.7765363128491626e-05,
+ "loss": 1.1832,
+ "step": 545
+ },
+ {
+ "epoch": 3.05,
+ "learning_rate": 4.748603351955307e-05,
+ "loss": 1.2019,
+ "step": 546
+ },
+ {
+ "epoch": 3.06,
+ "learning_rate": 4.7206703910614525e-05,
+ "loss": 1.2753,
+ "step": 547
+ },
+ {
+ "epoch": 3.06,
+ "learning_rate": 4.692737430167598e-05,
+ "loss": 1.2745,
+ "step": 548
+ },
+ {
+ "epoch": 3.07,
+ "learning_rate": 4.664804469273743e-05,
+ "loss": 1.3944,
+ "step": 549
+ },
+ {
+ "epoch": 3.07,
+ "learning_rate": 4.6368715083798884e-05,
+ "loss": 1.3676,
+ "step": 550
+ },
+ {
+ "epoch": 3.08,
+ "learning_rate": 4.6089385474860336e-05,
+ "loss": 1.3226,
+ "step": 551
+ },
+ {
+ "epoch": 3.08,
+ "learning_rate": 4.581005586592179e-05,
+ "loss": 1.134,
+ "step": 552
+ },
+ {
+ "epoch": 3.09,
+ "learning_rate": 4.553072625698324e-05,
+ "loss": 1.2143,
+ "step": 553
+ },
+ {
+ "epoch": 3.09,
+ "learning_rate": 4.5251396648044695e-05,
+ "loss": 1.2543,
+ "step": 554
+ },
+ {
+ "epoch": 3.1,
+ "learning_rate": 4.497206703910615e-05,
+ "loss": 1.2136,
+ "step": 555
+ },
+ {
+ "epoch": 3.11,
+ "learning_rate": 4.4692737430167594e-05,
+ "loss": 1.5303,
+ "step": 556
+ },
+ {
+ "epoch": 3.11,
+ "learning_rate": 4.4413407821229054e-05,
+ "loss": 1.2461,
+ "step": 557
+ },
+ {
+ "epoch": 3.12,
+ "learning_rate": 4.41340782122905e-05,
+ "loss": 1.3282,
+ "step": 558
+ },
+ {
+ "epoch": 3.12,
+ "learning_rate": 4.385474860335196e-05,
+ "loss": 1.201,
+ "step": 559
+ },
+ {
+ "epoch": 3.13,
+ "learning_rate": 4.3575418994413406e-05,
+ "loss": 1.188,
+ "step": 560
+ },
+ {
+ "epoch": 3.13,
+ "learning_rate": 4.3296089385474866e-05,
+ "loss": 1.2299,
+ "step": 561
+ },
+ {
+ "epoch": 3.14,
+ "learning_rate": 4.301675977653631e-05,
+ "loss": 1.2713,
+ "step": 562
+ },
+ {
+ "epoch": 3.15,
+ "learning_rate": 4.273743016759777e-05,
+ "loss": 1.2606,
+ "step": 563
+ },
+ {
+ "epoch": 3.15,
+ "learning_rate": 4.245810055865922e-05,
+ "loss": 1.2021,
+ "step": 564
+ },
+ {
+ "epoch": 3.16,
+ "learning_rate": 4.217877094972068e-05,
+ "loss": 1.243,
+ "step": 565
+ },
+ {
+ "epoch": 3.16,
+ "learning_rate": 4.1899441340782123e-05,
+ "loss": 1.2973,
+ "step": 566
+ },
+ {
+ "epoch": 3.17,
+ "learning_rate": 4.1620111731843576e-05,
+ "loss": 1.2369,
+ "step": 567
+ },
+ {
+ "epoch": 3.17,
+ "learning_rate": 4.134078212290503e-05,
+ "loss": 1.2729,
+ "step": 568
+ },
+ {
+ "epoch": 3.18,
+ "learning_rate": 4.106145251396648e-05,
+ "loss": 1.1511,
+ "step": 569
+ },
+ {
+ "epoch": 3.18,
+ "learning_rate": 4.0782122905027935e-05,
+ "loss": 1.3134,
+ "step": 570
+ },
+ {
+ "epoch": 3.19,
+ "learning_rate": 4.050279329608939e-05,
+ "loss": 1.3388,
+ "step": 571
+ },
+ {
+ "epoch": 3.2,
+ "learning_rate": 4.022346368715084e-05,
+ "loss": 1.3224,
+ "step": 572
+ },
+ {
+ "epoch": 3.2,
+ "learning_rate": 3.9944134078212294e-05,
+ "loss": 1.2827,
+ "step": 573
+ },
+ {
+ "epoch": 3.21,
+ "learning_rate": 3.966480446927375e-05,
+ "loss": 1.5265,
+ "step": 574
+ },
+ {
+ "epoch": 3.21,
+ "learning_rate": 3.93854748603352e-05,
+ "loss": 1.4008,
+ "step": 575
+ },
+ {
+ "epoch": 3.22,
+ "learning_rate": 3.9106145251396646e-05,
+ "loss": 1.1727,
+ "step": 576
+ },
+ {
+ "epoch": 3.22,
+ "learning_rate": 3.8826815642458106e-05,
+ "loss": 1.1999,
+ "step": 577
+ },
+ {
+ "epoch": 3.23,
+ "learning_rate": 3.854748603351955e-05,
+ "loss": 1.2031,
+ "step": 578
+ },
+ {
+ "epoch": 3.23,
+ "learning_rate": 3.826815642458101e-05,
+ "loss": 1.1469,
+ "step": 579
+ },
+ {
+ "epoch": 3.24,
+ "learning_rate": 3.798882681564246e-05,
+ "loss": 1.2784,
+ "step": 580
+ },
+ {
+ "epoch": 3.25,
+ "learning_rate": 3.770949720670392e-05,
+ "loss": 1.262,
+ "step": 581
+ },
+ {
+ "epoch": 3.25,
+ "learning_rate": 3.743016759776536e-05,
+ "loss": 1.224,
+ "step": 582
+ },
+ {
+ "epoch": 3.26,
+ "learning_rate": 3.715083798882682e-05,
+ "loss": 1.3554,
+ "step": 583
+ },
+ {
+ "epoch": 3.26,
+ "learning_rate": 3.687150837988827e-05,
+ "loss": 1.5137,
+ "step": 584
+ },
+ {
+ "epoch": 3.27,
+ "learning_rate": 3.659217877094972e-05,
+ "loss": 1.2462,
+ "step": 585
+ },
+ {
+ "epoch": 3.27,
+ "learning_rate": 3.6312849162011175e-05,
+ "loss": 1.2217,
+ "step": 586
+ },
+ {
+ "epoch": 3.28,
+ "learning_rate": 3.603351955307263e-05,
+ "loss": 1.2346,
+ "step": 587
+ },
+ {
+ "epoch": 3.28,
+ "learning_rate": 3.575418994413408e-05,
+ "loss": 1.3059,
+ "step": 588
+ },
+ {
+ "epoch": 3.29,
+ "learning_rate": 3.5474860335195534e-05,
+ "loss": 1.2355,
+ "step": 589
+ },
+ {
+ "epoch": 3.3,
+ "learning_rate": 3.519553072625699e-05,
+ "loss": 1.2233,
+ "step": 590
+ },
+ {
+ "epoch": 3.3,
+ "learning_rate": 3.491620111731844e-05,
+ "loss": 1.2014,
+ "step": 591
+ },
+ {
+ "epoch": 3.31,
+ "learning_rate": 3.463687150837989e-05,
+ "loss": 1.1919,
+ "step": 592
+ },
+ {
+ "epoch": 3.31,
+ "learning_rate": 3.4357541899441345e-05,
+ "loss": 1.2905,
+ "step": 593
+ },
+ {
+ "epoch": 3.32,
+ "learning_rate": 3.407821229050279e-05,
+ "loss": 1.0858,
+ "step": 594
+ },
+ {
+ "epoch": 3.32,
+ "learning_rate": 3.3798882681564244e-05,
+ "loss": 1.087,
+ "step": 595
+ },
+ {
+ "epoch": 3.33,
+ "learning_rate": 3.35195530726257e-05,
+ "loss": 1.1205,
+ "step": 596
+ },
+ {
+ "epoch": 3.34,
+ "learning_rate": 3.324022346368715e-05,
+ "loss": 1.1657,
+ "step": 597
+ },
+ {
+ "epoch": 3.34,
+ "learning_rate": 3.29608938547486e-05,
+ "loss": 1.5024,
+ "step": 598
+ },
+ {
+ "epoch": 3.35,
+ "learning_rate": 3.2681564245810056e-05,
+ "loss": 1.246,
+ "step": 599
+ },
+ {
+ "epoch": 3.35,
+ "learning_rate": 3.240223463687151e-05,
+ "loss": 1.4441,
+ "step": 600
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 716,
+ "num_train_epochs": 4,
+ "save_steps": 100,
+ "total_flos": 3.0851814594392064e+17,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-600/training_args.bin b/checkpoint-600/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8672c716e925d0028b4938db147703f58656ff7
--- /dev/null
+++ b/checkpoint-600/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:188ae1c421cc0c6435d1f71d8d3423ac4abc7dba0e6fc2efcbc4dbe77c741317
+size 4027
diff --git a/checkpoint-700/README.md b/checkpoint-700/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..08371015f02382e6fcba318f4aaea54ae52cd3c4
--- /dev/null
+++ b/checkpoint-700/README.md
@@ -0,0 +1,34 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: True
+- load_in_4bit: False
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: fp4
+- bnb_4bit_use_double_quant: False
+- bnb_4bit_compute_dtype: float32
+### Framework versions
+
+- PEFT 0.6.0.dev0
+
+- PEFT 0.6.0.dev0
diff --git a/checkpoint-700/adapter_config.json b/checkpoint-700/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c661f74e40394c079de68f07861d73c341803b67
--- /dev/null
+++ b/checkpoint-700/adapter_config.json
@@ -0,0 +1,25 @@
+{
+ "alpha_pattern": {},
+ "auto_mapping": null,
+ "base_model_name_or_path": "bigscience/bloomz-3b",
+ "bias": "none",
+ "fan_in_fan_out": false,
+ "inference_mode": true,
+ "init_lora_weights": true,
+ "layers_pattern": null,
+ "layers_to_transform": null,
+ "lora_alpha": 16,
+ "lora_dropout": 0.0,
+ "modules_to_save": null,
+ "peft_type": "LORA",
+ "r": 8,
+ "rank_pattern": {},
+ "revision": null,
+ "target_modules": [
+ "dense_4h_to_h",
+ "dense",
+ "dense_h_to_4h",
+ "query_key_value"
+ ],
+ "task_type": "CAUSAL_LM"
+}
\ No newline at end of file
diff --git a/checkpoint-700/adapter_model.bin b/checkpoint-700/adapter_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..0caec0282fb2027bbfc211849f9ba41039610438
--- /dev/null
+++ b/checkpoint-700/adapter_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d1b6626d0b27d106d068977d6a30f5fc34a5a358b55dafa50ff163993945015
+size 39409357
diff --git a/checkpoint-700/optimizer.pt b/checkpoint-700/optimizer.pt
new file mode 100644
index 0000000000000000000000000000000000000000..d65c53f0bec067303125e293857bdec3e9a6e893
--- /dev/null
+++ b/checkpoint-700/optimizer.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ae4fe25030e7252e0627af5e03d7f17a8d31b5da0e6730e68e43565c8cd8867
+size 78844421
diff --git a/checkpoint-700/rng_state.pth b/checkpoint-700/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..452f99ac534b9117d836494d73222e3d44e1523b
--- /dev/null
+++ b/checkpoint-700/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6869750f95a25c4e970298a33adf90e2d7ab52680bf3317239bff1b10103235
+size 14575
diff --git a/checkpoint-700/scheduler.pt b/checkpoint-700/scheduler.pt
new file mode 100644
index 0000000000000000000000000000000000000000..1358aafe4baa4670638ee88564b818147a3ea246
--- /dev/null
+++ b/checkpoint-700/scheduler.pt
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed66937ddb6bbdab2dab1a50e3279ef0e50829c8752c931e65a10f8a8aed77af
+size 627
diff --git a/checkpoint-700/special_tokens_map.json b/checkpoint-700/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdafe480f024ff444c7492147536765ce5d55a2d
--- /dev/null
+++ b/checkpoint-700/special_tokens_map.json
@@ -0,0 +1,6 @@
+{
+ "bos_token": "",
+ "eos_token": "",
+ "pad_token": "",
+ "unk_token": ""
+}
diff --git a/checkpoint-700/tokenizer.json b/checkpoint-700/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..673c31abdeadf6576c3c754df86459e1ad64e207
--- /dev/null
+++ b/checkpoint-700/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85b00d7db4df5df2e3f01cacc3feda246002a672f3356eec7f4b04a22eb0dfbe
+size 14500570
diff --git a/checkpoint-700/tokenizer_config.json b/checkpoint-700/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..4b56cc9c2965c07132c35df3e2972e93d98c82c3
--- /dev/null
+++ b/checkpoint-700/tokenizer_config.json
@@ -0,0 +1,10 @@
+{
+ "add_prefix_space": false,
+ "bos_token": "",
+ "clean_up_tokenization_spaces": false,
+ "eos_token": "",
+ "model_max_length": 1000000000000000019884624838656,
+ "pad_token": "",
+ "tokenizer_class": "BloomTokenizer",
+ "unk_token": ""
+}
diff --git a/checkpoint-700/trainer_state.json b/checkpoint-700/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..9168ec3620e1259528cd292a3d325413a2bd0a9a
--- /dev/null
+++ b/checkpoint-700/trainer_state.json
@@ -0,0 +1,4219 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 3.910614525139665,
+ "eval_steps": 500,
+ "global_step": 700,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019972067039106145,
+ "loss": 2.6443,
+ "step": 1
+ },
+ {
+ "epoch": 0.01,
+ "learning_rate": 0.00019944134078212292,
+ "loss": 2.4104,
+ "step": 2
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019916201117318435,
+ "loss": 2.4975,
+ "step": 3
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.00019888268156424582,
+ "loss": 2.3513,
+ "step": 4
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0001986033519553073,
+ "loss": 2.4274,
+ "step": 5
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.00019832402234636873,
+ "loss": 2.3628,
+ "step": 6
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.0001980446927374302,
+ "loss": 2.3567,
+ "step": 7
+ },
+ {
+ "epoch": 0.04,
+ "learning_rate": 0.00019776536312849163,
+ "loss": 2.4121,
+ "step": 8
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.00019748603351955307,
+ "loss": 2.4033,
+ "step": 9
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019720670391061454,
+ "loss": 2.2805,
+ "step": 10
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.00019692737430167598,
+ "loss": 2.2639,
+ "step": 11
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019664804469273744,
+ "loss": 2.2724,
+ "step": 12
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.00019636871508379888,
+ "loss": 2.332,
+ "step": 13
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019608938547486035,
+ "loss": 2.2261,
+ "step": 14
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.00019581005586592182,
+ "loss": 2.2208,
+ "step": 15
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.00019553072625698326,
+ "loss": 2.3351,
+ "step": 16
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0001952513966480447,
+ "loss": 2.2475,
+ "step": 17
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.00019497206703910616,
+ "loss": 2.3283,
+ "step": 18
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.0001946927374301676,
+ "loss": 2.1346,
+ "step": 19
+ },
+ {
+ "epoch": 0.11,
+ "learning_rate": 0.00019441340782122907,
+ "loss": 2.131,
+ "step": 20
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0001941340782122905,
+ "loss": 2.1718,
+ "step": 21
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.00019385474860335195,
+ "loss": 2.2446,
+ "step": 22
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0001935754189944134,
+ "loss": 2.306,
+ "step": 23
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.00019329608938547488,
+ "loss": 2.1908,
+ "step": 24
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.00019301675977653632,
+ "loss": 2.2844,
+ "step": 25
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019273743016759779,
+ "loss": 2.2235,
+ "step": 26
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.00019245810055865922,
+ "loss": 2.1842,
+ "step": 27
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019217877094972066,
+ "loss": 2.2675,
+ "step": 28
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.00019189944134078213,
+ "loss": 2.2532,
+ "step": 29
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019162011173184357,
+ "loss": 2.1788,
+ "step": 30
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.00019134078212290504,
+ "loss": 2.2494,
+ "step": 31
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.0001910614525139665,
+ "loss": 2.1995,
+ "step": 32
+ },
+ {
+ "epoch": 0.18,
+ "learning_rate": 0.00019078212290502794,
+ "loss": 2.1451,
+ "step": 33
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0001905027932960894,
+ "loss": 2.223,
+ "step": 34
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.00019022346368715085,
+ "loss": 2.2854,
+ "step": 35
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0001899441340782123,
+ "loss": 2.2265,
+ "step": 36
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.00018966480446927375,
+ "loss": 2.1214,
+ "step": 37
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0001893854748603352,
+ "loss": 2.1898,
+ "step": 38
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.00018910614525139666,
+ "loss": 2.1974,
+ "step": 39
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0001888268156424581,
+ "loss": 2.2259,
+ "step": 40
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018854748603351957,
+ "loss": 2.2094,
+ "step": 41
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.00018826815642458103,
+ "loss": 2.1731,
+ "step": 42
+ },
+ {
+ "epoch": 0.24,
+ "learning_rate": 0.00018798882681564247,
+ "loss": 2.2373,
+ "step": 43
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.0001877094972067039,
+ "loss": 2.2295,
+ "step": 44
+ },
+ {
+ "epoch": 0.25,
+ "learning_rate": 0.00018743016759776538,
+ "loss": 2.1947,
+ "step": 45
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018715083798882682,
+ "loss": 2.2115,
+ "step": 46
+ },
+ {
+ "epoch": 0.26,
+ "learning_rate": 0.00018687150837988828,
+ "loss": 2.1224,
+ "step": 47
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018659217877094972,
+ "loss": 2.2137,
+ "step": 48
+ },
+ {
+ "epoch": 0.27,
+ "learning_rate": 0.00018631284916201116,
+ "loss": 2.2338,
+ "step": 49
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.00018603351955307266,
+ "loss": 2.1298,
+ "step": 50
+ },
+ {
+ "epoch": 0.28,
+ "learning_rate": 0.0001857541899441341,
+ "loss": 2.0883,
+ "step": 51
+ },
+ {
+ "epoch": 0.29,
+ "learning_rate": 0.00018547486033519553,
+ "loss": 2.1216,
+ "step": 52
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.000185195530726257,
+ "loss": 2.2112,
+ "step": 53
+ },
+ {
+ "epoch": 0.3,
+ "learning_rate": 0.00018491620111731844,
+ "loss": 2.1224,
+ "step": 54
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.0001846368715083799,
+ "loss": 2.2375,
+ "step": 55
+ },
+ {
+ "epoch": 0.31,
+ "learning_rate": 0.00018435754189944135,
+ "loss": 2.2235,
+ "step": 56
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018407821229050279,
+ "loss": 2.1682,
+ "step": 57
+ },
+ {
+ "epoch": 0.32,
+ "learning_rate": 0.00018379888268156425,
+ "loss": 2.2077,
+ "step": 58
+ },
+ {
+ "epoch": 0.33,
+ "learning_rate": 0.00018351955307262572,
+ "loss": 2.1596,
+ "step": 59
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018324022346368716,
+ "loss": 2.1311,
+ "step": 60
+ },
+ {
+ "epoch": 0.34,
+ "learning_rate": 0.00018296089385474862,
+ "loss": 2.1333,
+ "step": 61
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018268156424581006,
+ "loss": 2.0901,
+ "step": 62
+ },
+ {
+ "epoch": 0.35,
+ "learning_rate": 0.00018240223463687153,
+ "loss": 2.1971,
+ "step": 63
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.00018212290502793297,
+ "loss": 2.2602,
+ "step": 64
+ },
+ {
+ "epoch": 0.36,
+ "learning_rate": 0.0001818435754189944,
+ "loss": 2.2194,
+ "step": 65
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018156424581005588,
+ "loss": 2.1218,
+ "step": 66
+ },
+ {
+ "epoch": 0.37,
+ "learning_rate": 0.00018128491620111731,
+ "loss": 2.2049,
+ "step": 67
+ },
+ {
+ "epoch": 0.38,
+ "learning_rate": 0.00018100558659217878,
+ "loss": 2.1521,
+ "step": 68
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.00018072625698324025,
+ "loss": 2.112,
+ "step": 69
+ },
+ {
+ "epoch": 0.39,
+ "learning_rate": 0.0001804469273743017,
+ "loss": 2.1906,
+ "step": 70
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.00018016759776536313,
+ "loss": 2.1717,
+ "step": 71
+ },
+ {
+ "epoch": 0.4,
+ "learning_rate": 0.0001798882681564246,
+ "loss": 2.0712,
+ "step": 72
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.00017960893854748603,
+ "loss": 2.141,
+ "step": 73
+ },
+ {
+ "epoch": 0.41,
+ "learning_rate": 0.0001793296089385475,
+ "loss": 2.0656,
+ "step": 74
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017905027932960894,
+ "loss": 2.1125,
+ "step": 75
+ },
+ {
+ "epoch": 0.42,
+ "learning_rate": 0.00017877094972067038,
+ "loss": 2.0869,
+ "step": 76
+ },
+ {
+ "epoch": 0.43,
+ "learning_rate": 0.00017849162011173187,
+ "loss": 2.2478,
+ "step": 77
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.0001782122905027933,
+ "loss": 2.1535,
+ "step": 78
+ },
+ {
+ "epoch": 0.44,
+ "learning_rate": 0.00017793296089385475,
+ "loss": 2.1927,
+ "step": 79
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017765363128491622,
+ "loss": 2.1213,
+ "step": 80
+ },
+ {
+ "epoch": 0.45,
+ "learning_rate": 0.00017737430167597766,
+ "loss": 2.0981,
+ "step": 81
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017709497206703912,
+ "loss": 2.1828,
+ "step": 82
+ },
+ {
+ "epoch": 0.46,
+ "learning_rate": 0.00017681564245810056,
+ "loss": 2.0562,
+ "step": 83
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.000176536312849162,
+ "loss": 2.1334,
+ "step": 84
+ },
+ {
+ "epoch": 0.47,
+ "learning_rate": 0.00017625698324022347,
+ "loss": 2.1225,
+ "step": 85
+ },
+ {
+ "epoch": 0.48,
+ "learning_rate": 0.00017597765363128493,
+ "loss": 2.2098,
+ "step": 86
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017569832402234637,
+ "loss": 2.1519,
+ "step": 87
+ },
+ {
+ "epoch": 0.49,
+ "learning_rate": 0.00017541899441340784,
+ "loss": 2.1132,
+ "step": 88
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017513966480446928,
+ "loss": 2.0333,
+ "step": 89
+ },
+ {
+ "epoch": 0.5,
+ "learning_rate": 0.00017486033519553075,
+ "loss": 2.2764,
+ "step": 90
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017458100558659218,
+ "loss": 2.1838,
+ "step": 91
+ },
+ {
+ "epoch": 0.51,
+ "learning_rate": 0.00017430167597765362,
+ "loss": 2.1386,
+ "step": 92
+ },
+ {
+ "epoch": 0.52,
+ "learning_rate": 0.0001740223463687151,
+ "loss": 2.1034,
+ "step": 93
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.00017374301675977656,
+ "loss": 2.0346,
+ "step": 94
+ },
+ {
+ "epoch": 0.53,
+ "learning_rate": 0.000173463687150838,
+ "loss": 2.0274,
+ "step": 95
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.00017318435754189946,
+ "loss": 2.1036,
+ "step": 96
+ },
+ {
+ "epoch": 0.54,
+ "learning_rate": 0.0001729050279329609,
+ "loss": 2.1208,
+ "step": 97
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.00017262569832402237,
+ "loss": 2.0572,
+ "step": 98
+ },
+ {
+ "epoch": 0.55,
+ "learning_rate": 0.0001723463687150838,
+ "loss": 2.1702,
+ "step": 99
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.00017206703910614525,
+ "loss": 2.1302,
+ "step": 100
+ },
+ {
+ "epoch": 0.56,
+ "learning_rate": 0.0001717877094972067,
+ "loss": 2.0175,
+ "step": 101
+ },
+ {
+ "epoch": 0.57,
+ "learning_rate": 0.00017150837988826815,
+ "loss": 2.1006,
+ "step": 102
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017122905027932962,
+ "loss": 2.0662,
+ "step": 103
+ },
+ {
+ "epoch": 0.58,
+ "learning_rate": 0.00017094972067039109,
+ "loss": 1.988,
+ "step": 104
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017067039106145253,
+ "loss": 2.1008,
+ "step": 105
+ },
+ {
+ "epoch": 0.59,
+ "learning_rate": 0.00017039106145251396,
+ "loss": 2.1482,
+ "step": 106
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00017011173184357543,
+ "loss": 2.1052,
+ "step": 107
+ },
+ {
+ "epoch": 0.6,
+ "learning_rate": 0.00016983240223463687,
+ "loss": 2.0978,
+ "step": 108
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016955307262569834,
+ "loss": 2.1303,
+ "step": 109
+ },
+ {
+ "epoch": 0.61,
+ "learning_rate": 0.00016927374301675978,
+ "loss": 2.0794,
+ "step": 110
+ },
+ {
+ "epoch": 0.62,
+ "learning_rate": 0.00016899441340782122,
+ "loss": 2.1059,
+ "step": 111
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.0001687150837988827,
+ "loss": 1.9642,
+ "step": 112
+ },
+ {
+ "epoch": 0.63,
+ "learning_rate": 0.00016843575418994415,
+ "loss": 2.0415,
+ "step": 113
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.0001681564245810056,
+ "loss": 2.0795,
+ "step": 114
+ },
+ {
+ "epoch": 0.64,
+ "learning_rate": 0.00016787709497206705,
+ "loss": 2.0238,
+ "step": 115
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.0001675977653631285,
+ "loss": 2.078,
+ "step": 116
+ },
+ {
+ "epoch": 0.65,
+ "learning_rate": 0.00016731843575418996,
+ "loss": 2.1362,
+ "step": 117
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.0001670391061452514,
+ "loss": 2.0552,
+ "step": 118
+ },
+ {
+ "epoch": 0.66,
+ "learning_rate": 0.00016675977653631284,
+ "loss": 2.213,
+ "step": 119
+ },
+ {
+ "epoch": 0.67,
+ "learning_rate": 0.0001664804469273743,
+ "loss": 1.9999,
+ "step": 120
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.00016620111731843577,
+ "loss": 2.06,
+ "step": 121
+ },
+ {
+ "epoch": 0.68,
+ "learning_rate": 0.0001659217877094972,
+ "loss": 2.0177,
+ "step": 122
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016564245810055868,
+ "loss": 2.0504,
+ "step": 123
+ },
+ {
+ "epoch": 0.69,
+ "learning_rate": 0.00016536312849162012,
+ "loss": 2.0585,
+ "step": 124
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016508379888268158,
+ "loss": 2.0273,
+ "step": 125
+ },
+ {
+ "epoch": 0.7,
+ "learning_rate": 0.00016480446927374302,
+ "loss": 2.0549,
+ "step": 126
+ },
+ {
+ "epoch": 0.71,
+ "learning_rate": 0.00016452513966480446,
+ "loss": 2.044,
+ "step": 127
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016424581005586593,
+ "loss": 2.0731,
+ "step": 128
+ },
+ {
+ "epoch": 0.72,
+ "learning_rate": 0.00016396648044692737,
+ "loss": 2.0568,
+ "step": 129
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.00016368715083798883,
+ "loss": 2.007,
+ "step": 130
+ },
+ {
+ "epoch": 0.73,
+ "learning_rate": 0.0001634078212290503,
+ "loss": 2.0707,
+ "step": 131
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.00016312849162011174,
+ "loss": 1.9793,
+ "step": 132
+ },
+ {
+ "epoch": 0.74,
+ "learning_rate": 0.0001628491620111732,
+ "loss": 2.1311,
+ "step": 133
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016256983240223465,
+ "loss": 2.0016,
+ "step": 134
+ },
+ {
+ "epoch": 0.75,
+ "learning_rate": 0.00016229050279329609,
+ "loss": 1.9945,
+ "step": 135
+ },
+ {
+ "epoch": 0.76,
+ "learning_rate": 0.00016201117318435755,
+ "loss": 2.0186,
+ "step": 136
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.000161731843575419,
+ "loss": 2.0971,
+ "step": 137
+ },
+ {
+ "epoch": 0.77,
+ "learning_rate": 0.00016145251396648046,
+ "loss": 2.0883,
+ "step": 138
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016117318435754192,
+ "loss": 2.0803,
+ "step": 139
+ },
+ {
+ "epoch": 0.78,
+ "learning_rate": 0.00016089385474860336,
+ "loss": 2.0617,
+ "step": 140
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016061452513966483,
+ "loss": 2.1265,
+ "step": 141
+ },
+ {
+ "epoch": 0.79,
+ "learning_rate": 0.00016033519553072627,
+ "loss": 2.0151,
+ "step": 142
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.0001600558659217877,
+ "loss": 1.996,
+ "step": 143
+ },
+ {
+ "epoch": 0.8,
+ "learning_rate": 0.00015977653631284918,
+ "loss": 2.0164,
+ "step": 144
+ },
+ {
+ "epoch": 0.81,
+ "learning_rate": 0.00015949720670391061,
+ "loss": 2.0314,
+ "step": 145
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015921787709497208,
+ "loss": 1.9501,
+ "step": 146
+ },
+ {
+ "epoch": 0.82,
+ "learning_rate": 0.00015893854748603352,
+ "loss": 2.087,
+ "step": 147
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.000158659217877095,
+ "loss": 2.0262,
+ "step": 148
+ },
+ {
+ "epoch": 0.83,
+ "learning_rate": 0.00015837988826815643,
+ "loss": 2.0765,
+ "step": 149
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.0001581005586592179,
+ "loss": 2.105,
+ "step": 150
+ },
+ {
+ "epoch": 0.84,
+ "learning_rate": 0.00015782122905027933,
+ "loss": 1.9863,
+ "step": 151
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.0001575418994413408,
+ "loss": 1.9873,
+ "step": 152
+ },
+ {
+ "epoch": 0.85,
+ "learning_rate": 0.00015726256983240224,
+ "loss": 2.0094,
+ "step": 153
+ },
+ {
+ "epoch": 0.86,
+ "learning_rate": 0.00015698324022346368,
+ "loss": 1.9141,
+ "step": 154
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015670391061452514,
+ "loss": 1.917,
+ "step": 155
+ },
+ {
+ "epoch": 0.87,
+ "learning_rate": 0.00015642458100558658,
+ "loss": 2.109,
+ "step": 156
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015614525139664805,
+ "loss": 1.9799,
+ "step": 157
+ },
+ {
+ "epoch": 0.88,
+ "learning_rate": 0.00015586592178770952,
+ "loss": 1.9571,
+ "step": 158
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015558659217877096,
+ "loss": 1.9931,
+ "step": 159
+ },
+ {
+ "epoch": 0.89,
+ "learning_rate": 0.00015530726256983242,
+ "loss": 2.1004,
+ "step": 160
+ },
+ {
+ "epoch": 0.9,
+ "learning_rate": 0.00015502793296089386,
+ "loss": 2.0385,
+ "step": 161
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.0001547486033519553,
+ "loss": 1.9751,
+ "step": 162
+ },
+ {
+ "epoch": 0.91,
+ "learning_rate": 0.00015446927374301677,
+ "loss": 2.0544,
+ "step": 163
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.0001541899441340782,
+ "loss": 2.0069,
+ "step": 164
+ },
+ {
+ "epoch": 0.92,
+ "learning_rate": 0.00015391061452513967,
+ "loss": 1.9576,
+ "step": 165
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015363128491620114,
+ "loss": 1.8991,
+ "step": 166
+ },
+ {
+ "epoch": 0.93,
+ "learning_rate": 0.00015335195530726258,
+ "loss": 1.9336,
+ "step": 167
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015307262569832405,
+ "loss": 1.9736,
+ "step": 168
+ },
+ {
+ "epoch": 0.94,
+ "learning_rate": 0.00015279329608938548,
+ "loss": 1.9702,
+ "step": 169
+ },
+ {
+ "epoch": 0.95,
+ "learning_rate": 0.00015251396648044692,
+ "loss": 1.9055,
+ "step": 170
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.0001522346368715084,
+ "loss": 2.0503,
+ "step": 171
+ },
+ {
+ "epoch": 0.96,
+ "learning_rate": 0.00015195530726256983,
+ "loss": 2.0039,
+ "step": 172
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.0001516759776536313,
+ "loss": 1.9406,
+ "step": 173
+ },
+ {
+ "epoch": 0.97,
+ "learning_rate": 0.00015139664804469274,
+ "loss": 2.0525,
+ "step": 174
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.0001511173184357542,
+ "loss": 1.9234,
+ "step": 175
+ },
+ {
+ "epoch": 0.98,
+ "learning_rate": 0.00015083798882681567,
+ "loss": 1.8614,
+ "step": 176
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.0001505586592178771,
+ "loss": 1.9616,
+ "step": 177
+ },
+ {
+ "epoch": 0.99,
+ "learning_rate": 0.00015027932960893855,
+ "loss": 1.9509,
+ "step": 178
+ },
+ {
+ "epoch": 1.0,
+ "learning_rate": 0.00015000000000000001,
+ "loss": 1.9592,
+ "step": 179
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014972067039106145,
+ "loss": 1.8991,
+ "step": 180
+ },
+ {
+ "epoch": 1.01,
+ "learning_rate": 0.00014944134078212292,
+ "loss": 1.9127,
+ "step": 181
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.00014916201117318436,
+ "loss": 1.8982,
+ "step": 182
+ },
+ {
+ "epoch": 1.02,
+ "learning_rate": 0.0001488826815642458,
+ "loss": 1.9534,
+ "step": 183
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.0001486033519553073,
+ "loss": 1.7794,
+ "step": 184
+ },
+ {
+ "epoch": 1.03,
+ "learning_rate": 0.00014832402234636873,
+ "loss": 1.7958,
+ "step": 185
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014804469273743017,
+ "loss": 1.8282,
+ "step": 186
+ },
+ {
+ "epoch": 1.04,
+ "learning_rate": 0.00014776536312849164,
+ "loss": 2.0423,
+ "step": 187
+ },
+ {
+ "epoch": 1.05,
+ "learning_rate": 0.00014748603351955308,
+ "loss": 1.9282,
+ "step": 188
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014720670391061454,
+ "loss": 1.9072,
+ "step": 189
+ },
+ {
+ "epoch": 1.06,
+ "learning_rate": 0.00014692737430167598,
+ "loss": 1.8665,
+ "step": 190
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.00014664804469273742,
+ "loss": 1.9021,
+ "step": 191
+ },
+ {
+ "epoch": 1.07,
+ "learning_rate": 0.0001463687150837989,
+ "loss": 1.7308,
+ "step": 192
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.00014608938547486035,
+ "loss": 1.9165,
+ "step": 193
+ },
+ {
+ "epoch": 1.08,
+ "learning_rate": 0.0001458100558659218,
+ "loss": 1.842,
+ "step": 194
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.00014553072625698326,
+ "loss": 1.9128,
+ "step": 195
+ },
+ {
+ "epoch": 1.09,
+ "learning_rate": 0.0001452513966480447,
+ "loss": 1.8005,
+ "step": 196
+ },
+ {
+ "epoch": 1.1,
+ "learning_rate": 0.00014497206703910614,
+ "loss": 1.8547,
+ "step": 197
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.0001446927374301676,
+ "loss": 1.9042,
+ "step": 198
+ },
+ {
+ "epoch": 1.11,
+ "learning_rate": 0.00014441340782122905,
+ "loss": 1.8609,
+ "step": 199
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.0001441340782122905,
+ "loss": 1.9591,
+ "step": 200
+ },
+ {
+ "epoch": 1.12,
+ "learning_rate": 0.00014385474860335195,
+ "loss": 1.8722,
+ "step": 201
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014357541899441342,
+ "loss": 1.8535,
+ "step": 202
+ },
+ {
+ "epoch": 1.13,
+ "learning_rate": 0.00014329608938547488,
+ "loss": 1.8676,
+ "step": 203
+ },
+ {
+ "epoch": 1.14,
+ "learning_rate": 0.00014301675977653632,
+ "loss": 1.8976,
+ "step": 204
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014273743016759776,
+ "loss": 1.7723,
+ "step": 205
+ },
+ {
+ "epoch": 1.15,
+ "learning_rate": 0.00014245810055865923,
+ "loss": 1.8165,
+ "step": 206
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014217877094972067,
+ "loss": 1.7811,
+ "step": 207
+ },
+ {
+ "epoch": 1.16,
+ "learning_rate": 0.00014189944134078214,
+ "loss": 1.908,
+ "step": 208
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014162011173184357,
+ "loss": 1.7663,
+ "step": 209
+ },
+ {
+ "epoch": 1.17,
+ "learning_rate": 0.00014134078212290501,
+ "loss": 1.6779,
+ "step": 210
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.0001410614525139665,
+ "loss": 1.9039,
+ "step": 211
+ },
+ {
+ "epoch": 1.18,
+ "learning_rate": 0.00014078212290502795,
+ "loss": 1.8033,
+ "step": 212
+ },
+ {
+ "epoch": 1.19,
+ "learning_rate": 0.00014050279329608939,
+ "loss": 1.8251,
+ "step": 213
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.00014022346368715085,
+ "loss": 1.8505,
+ "step": 214
+ },
+ {
+ "epoch": 1.2,
+ "learning_rate": 0.0001399441340782123,
+ "loss": 1.8147,
+ "step": 215
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.00013966480446927376,
+ "loss": 1.8419,
+ "step": 216
+ },
+ {
+ "epoch": 1.21,
+ "learning_rate": 0.0001393854748603352,
+ "loss": 1.8401,
+ "step": 217
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.00013910614525139664,
+ "loss": 1.8912,
+ "step": 218
+ },
+ {
+ "epoch": 1.22,
+ "learning_rate": 0.0001388268156424581,
+ "loss": 1.7548,
+ "step": 219
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.00013854748603351957,
+ "loss": 1.8741,
+ "step": 220
+ },
+ {
+ "epoch": 1.23,
+ "learning_rate": 0.000138268156424581,
+ "loss": 1.9549,
+ "step": 221
+ },
+ {
+ "epoch": 1.24,
+ "learning_rate": 0.00013798882681564248,
+ "loss": 1.9093,
+ "step": 222
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013770949720670392,
+ "loss": 1.7896,
+ "step": 223
+ },
+ {
+ "epoch": 1.25,
+ "learning_rate": 0.00013743016759776538,
+ "loss": 1.8491,
+ "step": 224
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013715083798882682,
+ "loss": 1.7851,
+ "step": 225
+ },
+ {
+ "epoch": 1.26,
+ "learning_rate": 0.00013687150837988826,
+ "loss": 1.6992,
+ "step": 226
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013659217877094973,
+ "loss": 1.9765,
+ "step": 227
+ },
+ {
+ "epoch": 1.27,
+ "learning_rate": 0.00013631284916201117,
+ "loss": 1.8179,
+ "step": 228
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.00013603351955307263,
+ "loss": 1.8548,
+ "step": 229
+ },
+ {
+ "epoch": 1.28,
+ "learning_rate": 0.0001357541899441341,
+ "loss": 1.8843,
+ "step": 230
+ },
+ {
+ "epoch": 1.29,
+ "learning_rate": 0.00013547486033519554,
+ "loss": 1.9105,
+ "step": 231
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013519553072625698,
+ "loss": 1.8748,
+ "step": 232
+ },
+ {
+ "epoch": 1.3,
+ "learning_rate": 0.00013491620111731844,
+ "loss": 1.7976,
+ "step": 233
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013463687150837988,
+ "loss": 1.7369,
+ "step": 234
+ },
+ {
+ "epoch": 1.31,
+ "learning_rate": 0.00013435754189944135,
+ "loss": 1.7808,
+ "step": 235
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.0001340782122905028,
+ "loss": 1.8385,
+ "step": 236
+ },
+ {
+ "epoch": 1.32,
+ "learning_rate": 0.00013379888268156423,
+ "loss": 1.8295,
+ "step": 237
+ },
+ {
+ "epoch": 1.33,
+ "learning_rate": 0.00013351955307262572,
+ "loss": 1.757,
+ "step": 238
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.00013324022346368716,
+ "loss": 1.7904,
+ "step": 239
+ },
+ {
+ "epoch": 1.34,
+ "learning_rate": 0.0001329608938547486,
+ "loss": 1.7632,
+ "step": 240
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.00013268156424581007,
+ "loss": 1.7867,
+ "step": 241
+ },
+ {
+ "epoch": 1.35,
+ "learning_rate": 0.0001324022346368715,
+ "loss": 1.8259,
+ "step": 242
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.00013212290502793297,
+ "loss": 1.6655,
+ "step": 243
+ },
+ {
+ "epoch": 1.36,
+ "learning_rate": 0.0001318435754189944,
+ "loss": 1.6848,
+ "step": 244
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013156424581005585,
+ "loss": 1.7931,
+ "step": 245
+ },
+ {
+ "epoch": 1.37,
+ "learning_rate": 0.00013128491620111732,
+ "loss": 1.7868,
+ "step": 246
+ },
+ {
+ "epoch": 1.38,
+ "learning_rate": 0.00013100558659217879,
+ "loss": 1.7732,
+ "step": 247
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.00013072625698324022,
+ "loss": 1.7851,
+ "step": 248
+ },
+ {
+ "epoch": 1.39,
+ "learning_rate": 0.0001304469273743017,
+ "loss": 1.7406,
+ "step": 249
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.00013016759776536313,
+ "loss": 1.5853,
+ "step": 250
+ },
+ {
+ "epoch": 1.4,
+ "learning_rate": 0.0001298882681564246,
+ "loss": 1.8271,
+ "step": 251
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012960893854748604,
+ "loss": 1.6054,
+ "step": 252
+ },
+ {
+ "epoch": 1.41,
+ "learning_rate": 0.00012932960893854748,
+ "loss": 1.6884,
+ "step": 253
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012905027932960894,
+ "loss": 1.7333,
+ "step": 254
+ },
+ {
+ "epoch": 1.42,
+ "learning_rate": 0.00012877094972067038,
+ "loss": 1.803,
+ "step": 255
+ },
+ {
+ "epoch": 1.43,
+ "learning_rate": 0.00012849162011173185,
+ "loss": 1.63,
+ "step": 256
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012821229050279331,
+ "loss": 1.8023,
+ "step": 257
+ },
+ {
+ "epoch": 1.44,
+ "learning_rate": 0.00012793296089385475,
+ "loss": 1.762,
+ "step": 258
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012765363128491622,
+ "loss": 1.7565,
+ "step": 259
+ },
+ {
+ "epoch": 1.45,
+ "learning_rate": 0.00012737430167597766,
+ "loss": 1.5937,
+ "step": 260
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.0001270949720670391,
+ "loss": 1.7208,
+ "step": 261
+ },
+ {
+ "epoch": 1.46,
+ "learning_rate": 0.00012681564245810057,
+ "loss": 1.8097,
+ "step": 262
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.000126536312849162,
+ "loss": 1.7042,
+ "step": 263
+ },
+ {
+ "epoch": 1.47,
+ "learning_rate": 0.00012625698324022347,
+ "loss": 1.7892,
+ "step": 264
+ },
+ {
+ "epoch": 1.48,
+ "learning_rate": 0.00012597765363128494,
+ "loss": 1.752,
+ "step": 265
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012569832402234638,
+ "loss": 1.7333,
+ "step": 266
+ },
+ {
+ "epoch": 1.49,
+ "learning_rate": 0.00012541899441340784,
+ "loss": 1.7492,
+ "step": 267
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012513966480446928,
+ "loss": 1.818,
+ "step": 268
+ },
+ {
+ "epoch": 1.5,
+ "learning_rate": 0.00012486033519553072,
+ "loss": 1.7146,
+ "step": 269
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.0001245810055865922,
+ "loss": 1.7958,
+ "step": 270
+ },
+ {
+ "epoch": 1.51,
+ "learning_rate": 0.00012430167597765363,
+ "loss": 1.7212,
+ "step": 271
+ },
+ {
+ "epoch": 1.52,
+ "learning_rate": 0.0001240223463687151,
+ "loss": 1.8835,
+ "step": 272
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.00012374301675977656,
+ "loss": 1.7486,
+ "step": 273
+ },
+ {
+ "epoch": 1.53,
+ "learning_rate": 0.000123463687150838,
+ "loss": 1.8103,
+ "step": 274
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.00012318435754189944,
+ "loss": 1.7557,
+ "step": 275
+ },
+ {
+ "epoch": 1.54,
+ "learning_rate": 0.0001229050279329609,
+ "loss": 1.6198,
+ "step": 276
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.00012262569832402235,
+ "loss": 1.6971,
+ "step": 277
+ },
+ {
+ "epoch": 1.55,
+ "learning_rate": 0.0001223463687150838,
+ "loss": 1.668,
+ "step": 278
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.00012206703910614525,
+ "loss": 1.8795,
+ "step": 279
+ },
+ {
+ "epoch": 1.56,
+ "learning_rate": 0.0001217877094972067,
+ "loss": 1.6412,
+ "step": 280
+ },
+ {
+ "epoch": 1.57,
+ "learning_rate": 0.00012150837988826816,
+ "loss": 1.7497,
+ "step": 281
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012122905027932962,
+ "loss": 1.5577,
+ "step": 282
+ },
+ {
+ "epoch": 1.58,
+ "learning_rate": 0.00012094972067039108,
+ "loss": 1.8049,
+ "step": 283
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012067039106145253,
+ "loss": 1.6834,
+ "step": 284
+ },
+ {
+ "epoch": 1.59,
+ "learning_rate": 0.00012039106145251397,
+ "loss": 1.7978,
+ "step": 285
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00012011173184357542,
+ "loss": 1.6558,
+ "step": 286
+ },
+ {
+ "epoch": 1.6,
+ "learning_rate": 0.00011983240223463687,
+ "loss": 1.7561,
+ "step": 287
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011955307262569833,
+ "loss": 1.6807,
+ "step": 288
+ },
+ {
+ "epoch": 1.61,
+ "learning_rate": 0.00011927374301675978,
+ "loss": 1.6937,
+ "step": 289
+ },
+ {
+ "epoch": 1.62,
+ "learning_rate": 0.00011899441340782122,
+ "loss": 1.6591,
+ "step": 290
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.0001187150837988827,
+ "loss": 1.6771,
+ "step": 291
+ },
+ {
+ "epoch": 1.63,
+ "learning_rate": 0.00011843575418994415,
+ "loss": 1.7743,
+ "step": 292
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011815642458100559,
+ "loss": 1.5857,
+ "step": 293
+ },
+ {
+ "epoch": 1.64,
+ "learning_rate": 0.00011787709497206705,
+ "loss": 1.6999,
+ "step": 294
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.0001175977653631285,
+ "loss": 1.5661,
+ "step": 295
+ },
+ {
+ "epoch": 1.65,
+ "learning_rate": 0.00011731843575418995,
+ "loss": 1.7235,
+ "step": 296
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.0001170391061452514,
+ "loss": 1.607,
+ "step": 297
+ },
+ {
+ "epoch": 1.66,
+ "learning_rate": 0.00011675977653631284,
+ "loss": 1.68,
+ "step": 298
+ },
+ {
+ "epoch": 1.67,
+ "learning_rate": 0.0001164804469273743,
+ "loss": 1.6938,
+ "step": 299
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011620111731843578,
+ "loss": 1.6315,
+ "step": 300
+ },
+ {
+ "epoch": 1.68,
+ "learning_rate": 0.00011592178770949722,
+ "loss": 1.6802,
+ "step": 301
+ },
+ {
+ "epoch": 1.69,
+ "learning_rate": 0.00011564245810055867,
+ "loss": 1.7174,
+ "step": 302
+ },
+ {
+ "epoch": 1.69,
+ "learning_rate": 0.00011536312849162012,
+ "loss": 1.5212,
+ "step": 303
+ },
+ {
+ "epoch": 1.7,
+ "learning_rate": 0.00011508379888268157,
+ "loss": 1.5808,
+ "step": 304
+ },
+ {
+ "epoch": 1.7,
+ "learning_rate": 0.00011480446927374303,
+ "loss": 1.6152,
+ "step": 305
+ },
+ {
+ "epoch": 1.71,
+ "learning_rate": 0.00011452513966480447,
+ "loss": 1.5435,
+ "step": 306
+ },
+ {
+ "epoch": 1.72,
+ "learning_rate": 0.00011424581005586592,
+ "loss": 1.6603,
+ "step": 307
+ },
+ {
+ "epoch": 1.72,
+ "learning_rate": 0.00011396648044692737,
+ "loss": 1.685,
+ "step": 308
+ },
+ {
+ "epoch": 1.73,
+ "learning_rate": 0.00011368715083798884,
+ "loss": 1.6002,
+ "step": 309
+ },
+ {
+ "epoch": 1.73,
+ "learning_rate": 0.00011340782122905029,
+ "loss": 1.6046,
+ "step": 310
+ },
+ {
+ "epoch": 1.74,
+ "learning_rate": 0.00011312849162011174,
+ "loss": 1.5969,
+ "step": 311
+ },
+ {
+ "epoch": 1.74,
+ "learning_rate": 0.0001128491620111732,
+ "loss": 1.5845,
+ "step": 312
+ },
+ {
+ "epoch": 1.75,
+ "learning_rate": 0.00011256983240223464,
+ "loss": 1.8183,
+ "step": 313
+ },
+ {
+ "epoch": 1.75,
+ "learning_rate": 0.00011229050279329609,
+ "loss": 1.6953,
+ "step": 314
+ },
+ {
+ "epoch": 1.76,
+ "learning_rate": 0.00011201117318435754,
+ "loss": 1.7787,
+ "step": 315
+ },
+ {
+ "epoch": 1.77,
+ "learning_rate": 0.000111731843575419,
+ "loss": 1.6422,
+ "step": 316
+ },
+ {
+ "epoch": 1.77,
+ "learning_rate": 0.00011145251396648045,
+ "loss": 1.7034,
+ "step": 317
+ },
+ {
+ "epoch": 1.78,
+ "learning_rate": 0.00011117318435754192,
+ "loss": 1.7301,
+ "step": 318
+ },
+ {
+ "epoch": 1.78,
+ "learning_rate": 0.00011089385474860337,
+ "loss": 1.7084,
+ "step": 319
+ },
+ {
+ "epoch": 1.79,
+ "learning_rate": 0.00011061452513966482,
+ "loss": 1.772,
+ "step": 320
+ },
+ {
+ "epoch": 1.79,
+ "learning_rate": 0.00011033519553072626,
+ "loss": 1.5733,
+ "step": 321
+ },
+ {
+ "epoch": 1.8,
+ "learning_rate": 0.00011005586592178771,
+ "loss": 1.6423,
+ "step": 322
+ },
+ {
+ "epoch": 1.8,
+ "learning_rate": 0.00010977653631284917,
+ "loss": 1.5809,
+ "step": 323
+ },
+ {
+ "epoch": 1.81,
+ "learning_rate": 0.00010949720670391062,
+ "loss": 1.6781,
+ "step": 324
+ },
+ {
+ "epoch": 1.82,
+ "learning_rate": 0.00010921787709497207,
+ "loss": 1.6788,
+ "step": 325
+ },
+ {
+ "epoch": 1.82,
+ "learning_rate": 0.00010893854748603351,
+ "loss": 1.6346,
+ "step": 326
+ },
+ {
+ "epoch": 1.83,
+ "learning_rate": 0.00010865921787709499,
+ "loss": 1.6634,
+ "step": 327
+ },
+ {
+ "epoch": 1.83,
+ "learning_rate": 0.00010837988826815643,
+ "loss": 1.7561,
+ "step": 328
+ },
+ {
+ "epoch": 1.84,
+ "learning_rate": 0.00010810055865921788,
+ "loss": 1.66,
+ "step": 329
+ },
+ {
+ "epoch": 1.84,
+ "learning_rate": 0.00010782122905027934,
+ "loss": 1.7298,
+ "step": 330
+ },
+ {
+ "epoch": 1.85,
+ "learning_rate": 0.00010754189944134079,
+ "loss": 1.6893,
+ "step": 331
+ },
+ {
+ "epoch": 1.85,
+ "learning_rate": 0.00010726256983240224,
+ "loss": 1.7631,
+ "step": 332
+ },
+ {
+ "epoch": 1.86,
+ "learning_rate": 0.00010698324022346368,
+ "loss": 1.6633,
+ "step": 333
+ },
+ {
+ "epoch": 1.87,
+ "learning_rate": 0.00010670391061452513,
+ "loss": 1.5388,
+ "step": 334
+ },
+ {
+ "epoch": 1.87,
+ "learning_rate": 0.00010642458100558659,
+ "loss": 1.6718,
+ "step": 335
+ },
+ {
+ "epoch": 1.88,
+ "learning_rate": 0.00010614525139664805,
+ "loss": 1.5536,
+ "step": 336
+ },
+ {
+ "epoch": 1.88,
+ "learning_rate": 0.00010586592178770951,
+ "loss": 1.6483,
+ "step": 337
+ },
+ {
+ "epoch": 1.89,
+ "learning_rate": 0.00010558659217877096,
+ "loss": 1.5774,
+ "step": 338
+ },
+ {
+ "epoch": 1.89,
+ "learning_rate": 0.00010530726256983241,
+ "loss": 1.6366,
+ "step": 339
+ },
+ {
+ "epoch": 1.9,
+ "learning_rate": 0.00010502793296089387,
+ "loss": 1.5567,
+ "step": 340
+ },
+ {
+ "epoch": 1.91,
+ "learning_rate": 0.0001047486033519553,
+ "loss": 1.5323,
+ "step": 341
+ },
+ {
+ "epoch": 1.91,
+ "learning_rate": 0.00010446927374301676,
+ "loss": 1.4608,
+ "step": 342
+ },
+ {
+ "epoch": 1.92,
+ "learning_rate": 0.00010418994413407821,
+ "loss": 1.5933,
+ "step": 343
+ },
+ {
+ "epoch": 1.92,
+ "learning_rate": 0.00010391061452513966,
+ "loss": 1.6625,
+ "step": 344
+ },
+ {
+ "epoch": 1.93,
+ "learning_rate": 0.00010363128491620113,
+ "loss": 1.7236,
+ "step": 345
+ },
+ {
+ "epoch": 1.93,
+ "learning_rate": 0.00010335195530726258,
+ "loss": 1.759,
+ "step": 346
+ },
+ {
+ "epoch": 1.94,
+ "learning_rate": 0.00010307262569832404,
+ "loss": 1.7248,
+ "step": 347
+ },
+ {
+ "epoch": 1.94,
+ "learning_rate": 0.00010279329608938548,
+ "loss": 1.5144,
+ "step": 348
+ },
+ {
+ "epoch": 1.95,
+ "learning_rate": 0.00010251396648044693,
+ "loss": 1.6905,
+ "step": 349
+ },
+ {
+ "epoch": 1.96,
+ "learning_rate": 0.00010223463687150838,
+ "loss": 1.6119,
+ "step": 350
+ },
+ {
+ "epoch": 1.96,
+ "learning_rate": 0.00010195530726256983,
+ "loss": 1.5464,
+ "step": 351
+ },
+ {
+ "epoch": 1.97,
+ "learning_rate": 0.00010167597765363129,
+ "loss": 1.6901,
+ "step": 352
+ },
+ {
+ "epoch": 1.97,
+ "learning_rate": 0.00010139664804469273,
+ "loss": 1.3511,
+ "step": 353
+ },
+ {
+ "epoch": 1.98,
+ "learning_rate": 0.0001011173184357542,
+ "loss": 1.5434,
+ "step": 354
+ },
+ {
+ "epoch": 1.98,
+ "learning_rate": 0.00010083798882681566,
+ "loss": 1.5891,
+ "step": 355
+ },
+ {
+ "epoch": 1.99,
+ "learning_rate": 0.0001005586592178771,
+ "loss": 1.6658,
+ "step": 356
+ },
+ {
+ "epoch": 1.99,
+ "learning_rate": 0.00010027932960893855,
+ "loss": 1.5657,
+ "step": 357
+ },
+ {
+ "epoch": 2.0,
+ "learning_rate": 0.0001,
+ "loss": 1.7005,
+ "step": 358
+ },
+ {
+ "epoch": 2.01,
+ "learning_rate": 9.972067039106146e-05,
+ "loss": 1.4202,
+ "step": 359
+ },
+ {
+ "epoch": 2.01,
+ "learning_rate": 9.944134078212291e-05,
+ "loss": 1.5262,
+ "step": 360
+ },
+ {
+ "epoch": 2.02,
+ "learning_rate": 9.916201117318436e-05,
+ "loss": 1.6323,
+ "step": 361
+ },
+ {
+ "epoch": 2.02,
+ "learning_rate": 9.888268156424582e-05,
+ "loss": 1.5521,
+ "step": 362
+ },
+ {
+ "epoch": 2.03,
+ "learning_rate": 9.860335195530727e-05,
+ "loss": 1.5762,
+ "step": 363
+ },
+ {
+ "epoch": 2.03,
+ "learning_rate": 9.832402234636872e-05,
+ "loss": 1.613,
+ "step": 364
+ },
+ {
+ "epoch": 2.04,
+ "learning_rate": 9.804469273743018e-05,
+ "loss": 1.4231,
+ "step": 365
+ },
+ {
+ "epoch": 2.04,
+ "learning_rate": 9.776536312849163e-05,
+ "loss": 1.5706,
+ "step": 366
+ },
+ {
+ "epoch": 2.05,
+ "learning_rate": 9.748603351955308e-05,
+ "loss": 1.5245,
+ "step": 367
+ },
+ {
+ "epoch": 2.06,
+ "learning_rate": 9.720670391061453e-05,
+ "loss": 1.4771,
+ "step": 368
+ },
+ {
+ "epoch": 2.06,
+ "learning_rate": 9.692737430167597e-05,
+ "loss": 1.596,
+ "step": 369
+ },
+ {
+ "epoch": 2.07,
+ "learning_rate": 9.664804469273744e-05,
+ "loss": 1.537,
+ "step": 370
+ },
+ {
+ "epoch": 2.07,
+ "learning_rate": 9.636871508379889e-05,
+ "loss": 1.4276,
+ "step": 371
+ },
+ {
+ "epoch": 2.08,
+ "learning_rate": 9.608938547486033e-05,
+ "loss": 1.4746,
+ "step": 372
+ },
+ {
+ "epoch": 2.08,
+ "learning_rate": 9.581005586592178e-05,
+ "loss": 1.4374,
+ "step": 373
+ },
+ {
+ "epoch": 2.09,
+ "learning_rate": 9.553072625698325e-05,
+ "loss": 1.4704,
+ "step": 374
+ },
+ {
+ "epoch": 2.09,
+ "learning_rate": 9.52513966480447e-05,
+ "loss": 1.5997,
+ "step": 375
+ },
+ {
+ "epoch": 2.1,
+ "learning_rate": 9.497206703910614e-05,
+ "loss": 1.5034,
+ "step": 376
+ },
+ {
+ "epoch": 2.11,
+ "learning_rate": 9.46927374301676e-05,
+ "loss": 1.6392,
+ "step": 377
+ },
+ {
+ "epoch": 2.11,
+ "learning_rate": 9.441340782122905e-05,
+ "loss": 1.5611,
+ "step": 378
+ },
+ {
+ "epoch": 2.12,
+ "learning_rate": 9.413407821229052e-05,
+ "loss": 1.3384,
+ "step": 379
+ },
+ {
+ "epoch": 2.12,
+ "learning_rate": 9.385474860335196e-05,
+ "loss": 1.5732,
+ "step": 380
+ },
+ {
+ "epoch": 2.13,
+ "learning_rate": 9.357541899441341e-05,
+ "loss": 1.3874,
+ "step": 381
+ },
+ {
+ "epoch": 2.13,
+ "learning_rate": 9.329608938547486e-05,
+ "loss": 1.4139,
+ "step": 382
+ },
+ {
+ "epoch": 2.14,
+ "learning_rate": 9.301675977653633e-05,
+ "loss": 1.5195,
+ "step": 383
+ },
+ {
+ "epoch": 2.15,
+ "learning_rate": 9.273743016759777e-05,
+ "loss": 1.4371,
+ "step": 384
+ },
+ {
+ "epoch": 2.15,
+ "learning_rate": 9.245810055865922e-05,
+ "loss": 1.4411,
+ "step": 385
+ },
+ {
+ "epoch": 2.16,
+ "learning_rate": 9.217877094972067e-05,
+ "loss": 1.5214,
+ "step": 386
+ },
+ {
+ "epoch": 2.16,
+ "learning_rate": 9.189944134078213e-05,
+ "loss": 1.4971,
+ "step": 387
+ },
+ {
+ "epoch": 2.17,
+ "learning_rate": 9.162011173184358e-05,
+ "loss": 1.3453,
+ "step": 388
+ },
+ {
+ "epoch": 2.17,
+ "learning_rate": 9.134078212290503e-05,
+ "loss": 1.394,
+ "step": 389
+ },
+ {
+ "epoch": 2.18,
+ "learning_rate": 9.106145251396648e-05,
+ "loss": 1.5058,
+ "step": 390
+ },
+ {
+ "epoch": 2.18,
+ "learning_rate": 9.078212290502794e-05,
+ "loss": 1.4855,
+ "step": 391
+ },
+ {
+ "epoch": 2.19,
+ "learning_rate": 9.050279329608939e-05,
+ "loss": 1.4647,
+ "step": 392
+ },
+ {
+ "epoch": 2.2,
+ "learning_rate": 9.022346368715084e-05,
+ "loss": 1.3435,
+ "step": 393
+ },
+ {
+ "epoch": 2.2,
+ "learning_rate": 8.99441340782123e-05,
+ "loss": 1.5815,
+ "step": 394
+ },
+ {
+ "epoch": 2.21,
+ "learning_rate": 8.966480446927375e-05,
+ "loss": 1.4742,
+ "step": 395
+ },
+ {
+ "epoch": 2.21,
+ "learning_rate": 8.938547486033519e-05,
+ "loss": 1.5389,
+ "step": 396
+ },
+ {
+ "epoch": 2.22,
+ "learning_rate": 8.910614525139666e-05,
+ "loss": 1.384,
+ "step": 397
+ },
+ {
+ "epoch": 2.22,
+ "learning_rate": 8.882681564245811e-05,
+ "loss": 1.3967,
+ "step": 398
+ },
+ {
+ "epoch": 2.23,
+ "learning_rate": 8.854748603351956e-05,
+ "loss": 1.442,
+ "step": 399
+ },
+ {
+ "epoch": 2.23,
+ "learning_rate": 8.8268156424581e-05,
+ "loss": 1.396,
+ "step": 400
+ },
+ {
+ "epoch": 2.24,
+ "learning_rate": 8.798882681564247e-05,
+ "loss": 1.4706,
+ "step": 401
+ },
+ {
+ "epoch": 2.25,
+ "learning_rate": 8.770949720670392e-05,
+ "loss": 1.4217,
+ "step": 402
+ },
+ {
+ "epoch": 2.25,
+ "learning_rate": 8.743016759776537e-05,
+ "loss": 1.5617,
+ "step": 403
+ },
+ {
+ "epoch": 2.26,
+ "learning_rate": 8.715083798882681e-05,
+ "loss": 1.5701,
+ "step": 404
+ },
+ {
+ "epoch": 2.26,
+ "learning_rate": 8.687150837988828e-05,
+ "loss": 1.3659,
+ "step": 405
+ },
+ {
+ "epoch": 2.27,
+ "learning_rate": 8.659217877094973e-05,
+ "loss": 1.3424,
+ "step": 406
+ },
+ {
+ "epoch": 2.27,
+ "learning_rate": 8.631284916201118e-05,
+ "loss": 1.5782,
+ "step": 407
+ },
+ {
+ "epoch": 2.28,
+ "learning_rate": 8.603351955307262e-05,
+ "loss": 1.3355,
+ "step": 408
+ },
+ {
+ "epoch": 2.28,
+ "learning_rate": 8.575418994413408e-05,
+ "loss": 1.4798,
+ "step": 409
+ },
+ {
+ "epoch": 2.29,
+ "learning_rate": 8.547486033519554e-05,
+ "loss": 1.3813,
+ "step": 410
+ },
+ {
+ "epoch": 2.3,
+ "learning_rate": 8.519553072625698e-05,
+ "loss": 1.6011,
+ "step": 411
+ },
+ {
+ "epoch": 2.3,
+ "learning_rate": 8.491620111731844e-05,
+ "loss": 1.4596,
+ "step": 412
+ },
+ {
+ "epoch": 2.31,
+ "learning_rate": 8.463687150837989e-05,
+ "loss": 1.4997,
+ "step": 413
+ },
+ {
+ "epoch": 2.31,
+ "learning_rate": 8.435754189944135e-05,
+ "loss": 1.5255,
+ "step": 414
+ },
+ {
+ "epoch": 2.32,
+ "learning_rate": 8.40782122905028e-05,
+ "loss": 1.5305,
+ "step": 415
+ },
+ {
+ "epoch": 2.32,
+ "learning_rate": 8.379888268156425e-05,
+ "loss": 1.3029,
+ "step": 416
+ },
+ {
+ "epoch": 2.33,
+ "learning_rate": 8.35195530726257e-05,
+ "loss": 1.2046,
+ "step": 417
+ },
+ {
+ "epoch": 2.34,
+ "learning_rate": 8.324022346368715e-05,
+ "loss": 1.4985,
+ "step": 418
+ },
+ {
+ "epoch": 2.34,
+ "learning_rate": 8.29608938547486e-05,
+ "loss": 1.3486,
+ "step": 419
+ },
+ {
+ "epoch": 2.35,
+ "learning_rate": 8.268156424581006e-05,
+ "loss": 1.2839,
+ "step": 420
+ },
+ {
+ "epoch": 2.35,
+ "learning_rate": 8.240223463687151e-05,
+ "loss": 1.4286,
+ "step": 421
+ },
+ {
+ "epoch": 2.36,
+ "learning_rate": 8.212290502793296e-05,
+ "loss": 1.3893,
+ "step": 422
+ },
+ {
+ "epoch": 2.36,
+ "learning_rate": 8.184357541899442e-05,
+ "loss": 1.5477,
+ "step": 423
+ },
+ {
+ "epoch": 2.37,
+ "learning_rate": 8.156424581005587e-05,
+ "loss": 1.4973,
+ "step": 424
+ },
+ {
+ "epoch": 2.37,
+ "learning_rate": 8.128491620111732e-05,
+ "loss": 1.3549,
+ "step": 425
+ },
+ {
+ "epoch": 2.38,
+ "learning_rate": 8.100558659217878e-05,
+ "loss": 1.3966,
+ "step": 426
+ },
+ {
+ "epoch": 2.39,
+ "learning_rate": 8.072625698324023e-05,
+ "loss": 1.3974,
+ "step": 427
+ },
+ {
+ "epoch": 2.39,
+ "learning_rate": 8.044692737430168e-05,
+ "loss": 1.5211,
+ "step": 428
+ },
+ {
+ "epoch": 2.4,
+ "learning_rate": 8.016759776536313e-05,
+ "loss": 1.4751,
+ "step": 429
+ },
+ {
+ "epoch": 2.4,
+ "learning_rate": 7.988826815642459e-05,
+ "loss": 1.4322,
+ "step": 430
+ },
+ {
+ "epoch": 2.41,
+ "learning_rate": 7.960893854748604e-05,
+ "loss": 1.5991,
+ "step": 431
+ },
+ {
+ "epoch": 2.41,
+ "learning_rate": 7.93296089385475e-05,
+ "loss": 1.3288,
+ "step": 432
+ },
+ {
+ "epoch": 2.42,
+ "learning_rate": 7.905027932960895e-05,
+ "loss": 1.4098,
+ "step": 433
+ },
+ {
+ "epoch": 2.42,
+ "learning_rate": 7.87709497206704e-05,
+ "loss": 1.4819,
+ "step": 434
+ },
+ {
+ "epoch": 2.43,
+ "learning_rate": 7.849162011173184e-05,
+ "loss": 1.4003,
+ "step": 435
+ },
+ {
+ "epoch": 2.44,
+ "learning_rate": 7.821229050279329e-05,
+ "loss": 1.3475,
+ "step": 436
+ },
+ {
+ "epoch": 2.44,
+ "learning_rate": 7.793296089385476e-05,
+ "loss": 1.3354,
+ "step": 437
+ },
+ {
+ "epoch": 2.45,
+ "learning_rate": 7.765363128491621e-05,
+ "loss": 1.3144,
+ "step": 438
+ },
+ {
+ "epoch": 2.45,
+ "learning_rate": 7.737430167597765e-05,
+ "loss": 1.4118,
+ "step": 439
+ },
+ {
+ "epoch": 2.46,
+ "learning_rate": 7.70949720670391e-05,
+ "loss": 1.3716,
+ "step": 440
+ },
+ {
+ "epoch": 2.46,
+ "learning_rate": 7.681564245810057e-05,
+ "loss": 1.4191,
+ "step": 441
+ },
+ {
+ "epoch": 2.47,
+ "learning_rate": 7.653631284916202e-05,
+ "loss": 1.5173,
+ "step": 442
+ },
+ {
+ "epoch": 2.47,
+ "learning_rate": 7.625698324022346e-05,
+ "loss": 1.3005,
+ "step": 443
+ },
+ {
+ "epoch": 2.48,
+ "learning_rate": 7.597765363128491e-05,
+ "loss": 1.4586,
+ "step": 444
+ },
+ {
+ "epoch": 2.49,
+ "learning_rate": 7.569832402234637e-05,
+ "loss": 1.5086,
+ "step": 445
+ },
+ {
+ "epoch": 2.49,
+ "learning_rate": 7.541899441340783e-05,
+ "loss": 1.5446,
+ "step": 446
+ },
+ {
+ "epoch": 2.5,
+ "learning_rate": 7.513966480446927e-05,
+ "loss": 1.3274,
+ "step": 447
+ },
+ {
+ "epoch": 2.5,
+ "learning_rate": 7.486033519553073e-05,
+ "loss": 1.4424,
+ "step": 448
+ },
+ {
+ "epoch": 2.51,
+ "learning_rate": 7.458100558659218e-05,
+ "loss": 1.2836,
+ "step": 449
+ },
+ {
+ "epoch": 2.51,
+ "learning_rate": 7.430167597765365e-05,
+ "loss": 1.4322,
+ "step": 450
+ },
+ {
+ "epoch": 2.52,
+ "learning_rate": 7.402234636871509e-05,
+ "loss": 1.4168,
+ "step": 451
+ },
+ {
+ "epoch": 2.53,
+ "learning_rate": 7.374301675977654e-05,
+ "loss": 1.3892,
+ "step": 452
+ },
+ {
+ "epoch": 2.53,
+ "learning_rate": 7.346368715083799e-05,
+ "loss": 1.6538,
+ "step": 453
+ },
+ {
+ "epoch": 2.54,
+ "learning_rate": 7.318435754189944e-05,
+ "loss": 1.4573,
+ "step": 454
+ },
+ {
+ "epoch": 2.54,
+ "learning_rate": 7.29050279329609e-05,
+ "loss": 1.4593,
+ "step": 455
+ },
+ {
+ "epoch": 2.55,
+ "learning_rate": 7.262569832402235e-05,
+ "loss": 1.4316,
+ "step": 456
+ },
+ {
+ "epoch": 2.55,
+ "learning_rate": 7.23463687150838e-05,
+ "loss": 1.2702,
+ "step": 457
+ },
+ {
+ "epoch": 2.56,
+ "learning_rate": 7.206703910614526e-05,
+ "loss": 1.5458,
+ "step": 458
+ },
+ {
+ "epoch": 2.56,
+ "learning_rate": 7.178770949720671e-05,
+ "loss": 1.5179,
+ "step": 459
+ },
+ {
+ "epoch": 2.57,
+ "learning_rate": 7.150837988826816e-05,
+ "loss": 1.3844,
+ "step": 460
+ },
+ {
+ "epoch": 2.58,
+ "learning_rate": 7.122905027932961e-05,
+ "loss": 1.3698,
+ "step": 461
+ },
+ {
+ "epoch": 2.58,
+ "learning_rate": 7.094972067039107e-05,
+ "loss": 1.4396,
+ "step": 462
+ },
+ {
+ "epoch": 2.59,
+ "learning_rate": 7.067039106145251e-05,
+ "loss": 1.5486,
+ "step": 463
+ },
+ {
+ "epoch": 2.59,
+ "learning_rate": 7.039106145251397e-05,
+ "loss": 1.2032,
+ "step": 464
+ },
+ {
+ "epoch": 2.6,
+ "learning_rate": 7.011173184357543e-05,
+ "loss": 1.3434,
+ "step": 465
+ },
+ {
+ "epoch": 2.6,
+ "learning_rate": 6.983240223463688e-05,
+ "loss": 1.4065,
+ "step": 466
+ },
+ {
+ "epoch": 2.61,
+ "learning_rate": 6.955307262569832e-05,
+ "loss": 1.3362,
+ "step": 467
+ },
+ {
+ "epoch": 2.61,
+ "learning_rate": 6.927374301675979e-05,
+ "loss": 1.375,
+ "step": 468
+ },
+ {
+ "epoch": 2.62,
+ "learning_rate": 6.899441340782124e-05,
+ "loss": 1.4549,
+ "step": 469
+ },
+ {
+ "epoch": 2.63,
+ "learning_rate": 6.871508379888269e-05,
+ "loss": 1.3724,
+ "step": 470
+ },
+ {
+ "epoch": 2.63,
+ "learning_rate": 6.843575418994413e-05,
+ "loss": 1.1602,
+ "step": 471
+ },
+ {
+ "epoch": 2.64,
+ "learning_rate": 6.815642458100558e-05,
+ "loss": 1.4346,
+ "step": 472
+ },
+ {
+ "epoch": 2.64,
+ "learning_rate": 6.787709497206705e-05,
+ "loss": 1.2877,
+ "step": 473
+ },
+ {
+ "epoch": 2.65,
+ "learning_rate": 6.759776536312849e-05,
+ "loss": 1.394,
+ "step": 474
+ },
+ {
+ "epoch": 2.65,
+ "learning_rate": 6.731843575418994e-05,
+ "loss": 1.3387,
+ "step": 475
+ },
+ {
+ "epoch": 2.66,
+ "learning_rate": 6.70391061452514e-05,
+ "loss": 1.41,
+ "step": 476
+ },
+ {
+ "epoch": 2.66,
+ "learning_rate": 6.675977653631286e-05,
+ "loss": 1.4299,
+ "step": 477
+ },
+ {
+ "epoch": 2.67,
+ "learning_rate": 6.64804469273743e-05,
+ "loss": 1.3824,
+ "step": 478
+ },
+ {
+ "epoch": 2.68,
+ "learning_rate": 6.620111731843575e-05,
+ "loss": 1.2656,
+ "step": 479
+ },
+ {
+ "epoch": 2.68,
+ "learning_rate": 6.59217877094972e-05,
+ "loss": 1.2811,
+ "step": 480
+ },
+ {
+ "epoch": 2.69,
+ "learning_rate": 6.564245810055866e-05,
+ "loss": 1.6839,
+ "step": 481
+ },
+ {
+ "epoch": 2.69,
+ "learning_rate": 6.536312849162011e-05,
+ "loss": 1.3089,
+ "step": 482
+ },
+ {
+ "epoch": 2.7,
+ "learning_rate": 6.508379888268157e-05,
+ "loss": 1.4323,
+ "step": 483
+ },
+ {
+ "epoch": 2.7,
+ "learning_rate": 6.480446927374302e-05,
+ "loss": 1.3439,
+ "step": 484
+ },
+ {
+ "epoch": 2.71,
+ "learning_rate": 6.452513966480447e-05,
+ "loss": 1.3614,
+ "step": 485
+ },
+ {
+ "epoch": 2.72,
+ "learning_rate": 6.424581005586592e-05,
+ "loss": 1.4974,
+ "step": 486
+ },
+ {
+ "epoch": 2.72,
+ "learning_rate": 6.396648044692738e-05,
+ "loss": 1.4963,
+ "step": 487
+ },
+ {
+ "epoch": 2.73,
+ "learning_rate": 6.368715083798883e-05,
+ "loss": 1.2902,
+ "step": 488
+ },
+ {
+ "epoch": 2.73,
+ "learning_rate": 6.340782122905028e-05,
+ "loss": 1.3226,
+ "step": 489
+ },
+ {
+ "epoch": 2.74,
+ "learning_rate": 6.312849162011174e-05,
+ "loss": 1.4353,
+ "step": 490
+ },
+ {
+ "epoch": 2.74,
+ "learning_rate": 6.284916201117319e-05,
+ "loss": 1.281,
+ "step": 491
+ },
+ {
+ "epoch": 2.75,
+ "learning_rate": 6.256983240223464e-05,
+ "loss": 1.3755,
+ "step": 492
+ },
+ {
+ "epoch": 2.75,
+ "learning_rate": 6.22905027932961e-05,
+ "loss": 1.3811,
+ "step": 493
+ },
+ {
+ "epoch": 2.76,
+ "learning_rate": 6.201117318435755e-05,
+ "loss": 1.4627,
+ "step": 494
+ },
+ {
+ "epoch": 2.77,
+ "learning_rate": 6.1731843575419e-05,
+ "loss": 1.3482,
+ "step": 495
+ },
+ {
+ "epoch": 2.77,
+ "learning_rate": 6.145251396648045e-05,
+ "loss": 1.4274,
+ "step": 496
+ },
+ {
+ "epoch": 2.78,
+ "learning_rate": 6.11731843575419e-05,
+ "loss": 1.3465,
+ "step": 497
+ },
+ {
+ "epoch": 2.78,
+ "learning_rate": 6.089385474860335e-05,
+ "loss": 1.2319,
+ "step": 498
+ },
+ {
+ "epoch": 2.79,
+ "learning_rate": 6.061452513966481e-05,
+ "loss": 1.2835,
+ "step": 499
+ },
+ {
+ "epoch": 2.79,
+ "learning_rate": 6.0335195530726265e-05,
+ "loss": 1.4836,
+ "step": 500
+ },
+ {
+ "epoch": 2.8,
+ "learning_rate": 6.005586592178771e-05,
+ "loss": 1.3878,
+ "step": 501
+ },
+ {
+ "epoch": 2.8,
+ "learning_rate": 5.9776536312849164e-05,
+ "loss": 1.4367,
+ "step": 502
+ },
+ {
+ "epoch": 2.81,
+ "learning_rate": 5.949720670391061e-05,
+ "loss": 1.2994,
+ "step": 503
+ },
+ {
+ "epoch": 2.82,
+ "learning_rate": 5.9217877094972076e-05,
+ "loss": 1.3599,
+ "step": 504
+ },
+ {
+ "epoch": 2.82,
+ "learning_rate": 5.893854748603352e-05,
+ "loss": 1.5357,
+ "step": 505
+ },
+ {
+ "epoch": 2.83,
+ "learning_rate": 5.8659217877094976e-05,
+ "loss": 1.4497,
+ "step": 506
+ },
+ {
+ "epoch": 2.83,
+ "learning_rate": 5.837988826815642e-05,
+ "loss": 1.3496,
+ "step": 507
+ },
+ {
+ "epoch": 2.84,
+ "learning_rate": 5.810055865921789e-05,
+ "loss": 1.294,
+ "step": 508
+ },
+ {
+ "epoch": 2.84,
+ "learning_rate": 5.7821229050279334e-05,
+ "loss": 1.4204,
+ "step": 509
+ },
+ {
+ "epoch": 2.85,
+ "learning_rate": 5.754189944134079e-05,
+ "loss": 1.5231,
+ "step": 510
+ },
+ {
+ "epoch": 2.85,
+ "learning_rate": 5.726256983240223e-05,
+ "loss": 1.3165,
+ "step": 511
+ },
+ {
+ "epoch": 2.86,
+ "learning_rate": 5.6983240223463686e-05,
+ "loss": 1.0796,
+ "step": 512
+ },
+ {
+ "epoch": 2.87,
+ "learning_rate": 5.6703910614525146e-05,
+ "loss": 1.4086,
+ "step": 513
+ },
+ {
+ "epoch": 2.87,
+ "learning_rate": 5.64245810055866e-05,
+ "loss": 1.2895,
+ "step": 514
+ },
+ {
+ "epoch": 2.88,
+ "learning_rate": 5.6145251396648045e-05,
+ "loss": 1.3024,
+ "step": 515
+ },
+ {
+ "epoch": 2.88,
+ "learning_rate": 5.58659217877095e-05,
+ "loss": 1.2479,
+ "step": 516
+ },
+ {
+ "epoch": 2.89,
+ "learning_rate": 5.558659217877096e-05,
+ "loss": 1.5232,
+ "step": 517
+ },
+ {
+ "epoch": 2.89,
+ "learning_rate": 5.530726256983241e-05,
+ "loss": 1.2654,
+ "step": 518
+ },
+ {
+ "epoch": 2.9,
+ "learning_rate": 5.502793296089386e-05,
+ "loss": 1.3018,
+ "step": 519
+ },
+ {
+ "epoch": 2.91,
+ "learning_rate": 5.474860335195531e-05,
+ "loss": 1.3077,
+ "step": 520
+ },
+ {
+ "epoch": 2.91,
+ "learning_rate": 5.4469273743016756e-05,
+ "loss": 1.3501,
+ "step": 521
+ },
+ {
+ "epoch": 2.92,
+ "learning_rate": 5.4189944134078215e-05,
+ "loss": 1.2607,
+ "step": 522
+ },
+ {
+ "epoch": 2.92,
+ "learning_rate": 5.391061452513967e-05,
+ "loss": 1.3819,
+ "step": 523
+ },
+ {
+ "epoch": 2.93,
+ "learning_rate": 5.363128491620112e-05,
+ "loss": 1.321,
+ "step": 524
+ },
+ {
+ "epoch": 2.93,
+ "learning_rate": 5.335195530726257e-05,
+ "loss": 1.3351,
+ "step": 525
+ },
+ {
+ "epoch": 2.94,
+ "learning_rate": 5.307262569832403e-05,
+ "loss": 1.2205,
+ "step": 526
+ },
+ {
+ "epoch": 2.94,
+ "learning_rate": 5.279329608938548e-05,
+ "loss": 1.2579,
+ "step": 527
+ },
+ {
+ "epoch": 2.95,
+ "learning_rate": 5.251396648044693e-05,
+ "loss": 1.2771,
+ "step": 528
+ },
+ {
+ "epoch": 2.96,
+ "learning_rate": 5.223463687150838e-05,
+ "loss": 1.3492,
+ "step": 529
+ },
+ {
+ "epoch": 2.96,
+ "learning_rate": 5.195530726256983e-05,
+ "loss": 1.3232,
+ "step": 530
+ },
+ {
+ "epoch": 2.97,
+ "learning_rate": 5.167597765363129e-05,
+ "loss": 1.2223,
+ "step": 531
+ },
+ {
+ "epoch": 2.97,
+ "learning_rate": 5.139664804469274e-05,
+ "loss": 1.3983,
+ "step": 532
+ },
+ {
+ "epoch": 2.98,
+ "learning_rate": 5.111731843575419e-05,
+ "loss": 1.2987,
+ "step": 533
+ },
+ {
+ "epoch": 2.98,
+ "learning_rate": 5.0837988826815644e-05,
+ "loss": 1.3717,
+ "step": 534
+ },
+ {
+ "epoch": 2.99,
+ "learning_rate": 5.05586592178771e-05,
+ "loss": 1.4467,
+ "step": 535
+ },
+ {
+ "epoch": 2.99,
+ "learning_rate": 5.027932960893855e-05,
+ "loss": 1.4314,
+ "step": 536
+ },
+ {
+ "epoch": 3.0,
+ "learning_rate": 5e-05,
+ "loss": 1.6593,
+ "step": 537
+ },
+ {
+ "epoch": 3.01,
+ "learning_rate": 4.9720670391061455e-05,
+ "loss": 1.3425,
+ "step": 538
+ },
+ {
+ "epoch": 3.01,
+ "learning_rate": 4.944134078212291e-05,
+ "loss": 1.4171,
+ "step": 539
+ },
+ {
+ "epoch": 3.02,
+ "learning_rate": 4.916201117318436e-05,
+ "loss": 1.3766,
+ "step": 540
+ },
+ {
+ "epoch": 3.02,
+ "learning_rate": 4.8882681564245814e-05,
+ "loss": 1.1605,
+ "step": 541
+ },
+ {
+ "epoch": 3.03,
+ "learning_rate": 4.860335195530727e-05,
+ "loss": 1.3132,
+ "step": 542
+ },
+ {
+ "epoch": 3.03,
+ "learning_rate": 4.832402234636872e-05,
+ "loss": 1.4128,
+ "step": 543
+ },
+ {
+ "epoch": 3.04,
+ "learning_rate": 4.8044692737430166e-05,
+ "loss": 1.4902,
+ "step": 544
+ },
+ {
+ "epoch": 3.04,
+ "learning_rate": 4.7765363128491626e-05,
+ "loss": 1.1832,
+ "step": 545
+ },
+ {
+ "epoch": 3.05,
+ "learning_rate": 4.748603351955307e-05,
+ "loss": 1.2019,
+ "step": 546
+ },
+ {
+ "epoch": 3.06,
+ "learning_rate": 4.7206703910614525e-05,
+ "loss": 1.2753,
+ "step": 547
+ },
+ {
+ "epoch": 3.06,
+ "learning_rate": 4.692737430167598e-05,
+ "loss": 1.2745,
+ "step": 548
+ },
+ {
+ "epoch": 3.07,
+ "learning_rate": 4.664804469273743e-05,
+ "loss": 1.3944,
+ "step": 549
+ },
+ {
+ "epoch": 3.07,
+ "learning_rate": 4.6368715083798884e-05,
+ "loss": 1.3676,
+ "step": 550
+ },
+ {
+ "epoch": 3.08,
+ "learning_rate": 4.6089385474860336e-05,
+ "loss": 1.3226,
+ "step": 551
+ },
+ {
+ "epoch": 3.08,
+ "learning_rate": 4.581005586592179e-05,
+ "loss": 1.134,
+ "step": 552
+ },
+ {
+ "epoch": 3.09,
+ "learning_rate": 4.553072625698324e-05,
+ "loss": 1.2143,
+ "step": 553
+ },
+ {
+ "epoch": 3.09,
+ "learning_rate": 4.5251396648044695e-05,
+ "loss": 1.2543,
+ "step": 554
+ },
+ {
+ "epoch": 3.1,
+ "learning_rate": 4.497206703910615e-05,
+ "loss": 1.2136,
+ "step": 555
+ },
+ {
+ "epoch": 3.11,
+ "learning_rate": 4.4692737430167594e-05,
+ "loss": 1.5303,
+ "step": 556
+ },
+ {
+ "epoch": 3.11,
+ "learning_rate": 4.4413407821229054e-05,
+ "loss": 1.2461,
+ "step": 557
+ },
+ {
+ "epoch": 3.12,
+ "learning_rate": 4.41340782122905e-05,
+ "loss": 1.3282,
+ "step": 558
+ },
+ {
+ "epoch": 3.12,
+ "learning_rate": 4.385474860335196e-05,
+ "loss": 1.201,
+ "step": 559
+ },
+ {
+ "epoch": 3.13,
+ "learning_rate": 4.3575418994413406e-05,
+ "loss": 1.188,
+ "step": 560
+ },
+ {
+ "epoch": 3.13,
+ "learning_rate": 4.3296089385474866e-05,
+ "loss": 1.2299,
+ "step": 561
+ },
+ {
+ "epoch": 3.14,
+ "learning_rate": 4.301675977653631e-05,
+ "loss": 1.2713,
+ "step": 562
+ },
+ {
+ "epoch": 3.15,
+ "learning_rate": 4.273743016759777e-05,
+ "loss": 1.2606,
+ "step": 563
+ },
+ {
+ "epoch": 3.15,
+ "learning_rate": 4.245810055865922e-05,
+ "loss": 1.2021,
+ "step": 564
+ },
+ {
+ "epoch": 3.16,
+ "learning_rate": 4.217877094972068e-05,
+ "loss": 1.243,
+ "step": 565
+ },
+ {
+ "epoch": 3.16,
+ "learning_rate": 4.1899441340782123e-05,
+ "loss": 1.2973,
+ "step": 566
+ },
+ {
+ "epoch": 3.17,
+ "learning_rate": 4.1620111731843576e-05,
+ "loss": 1.2369,
+ "step": 567
+ },
+ {
+ "epoch": 3.17,
+ "learning_rate": 4.134078212290503e-05,
+ "loss": 1.2729,
+ "step": 568
+ },
+ {
+ "epoch": 3.18,
+ "learning_rate": 4.106145251396648e-05,
+ "loss": 1.1511,
+ "step": 569
+ },
+ {
+ "epoch": 3.18,
+ "learning_rate": 4.0782122905027935e-05,
+ "loss": 1.3134,
+ "step": 570
+ },
+ {
+ "epoch": 3.19,
+ "learning_rate": 4.050279329608939e-05,
+ "loss": 1.3388,
+ "step": 571
+ },
+ {
+ "epoch": 3.2,
+ "learning_rate": 4.022346368715084e-05,
+ "loss": 1.3224,
+ "step": 572
+ },
+ {
+ "epoch": 3.2,
+ "learning_rate": 3.9944134078212294e-05,
+ "loss": 1.2827,
+ "step": 573
+ },
+ {
+ "epoch": 3.21,
+ "learning_rate": 3.966480446927375e-05,
+ "loss": 1.5265,
+ "step": 574
+ },
+ {
+ "epoch": 3.21,
+ "learning_rate": 3.93854748603352e-05,
+ "loss": 1.4008,
+ "step": 575
+ },
+ {
+ "epoch": 3.22,
+ "learning_rate": 3.9106145251396646e-05,
+ "loss": 1.1727,
+ "step": 576
+ },
+ {
+ "epoch": 3.22,
+ "learning_rate": 3.8826815642458106e-05,
+ "loss": 1.1999,
+ "step": 577
+ },
+ {
+ "epoch": 3.23,
+ "learning_rate": 3.854748603351955e-05,
+ "loss": 1.2031,
+ "step": 578
+ },
+ {
+ "epoch": 3.23,
+ "learning_rate": 3.826815642458101e-05,
+ "loss": 1.1469,
+ "step": 579
+ },
+ {
+ "epoch": 3.24,
+ "learning_rate": 3.798882681564246e-05,
+ "loss": 1.2784,
+ "step": 580
+ },
+ {
+ "epoch": 3.25,
+ "learning_rate": 3.770949720670392e-05,
+ "loss": 1.262,
+ "step": 581
+ },
+ {
+ "epoch": 3.25,
+ "learning_rate": 3.743016759776536e-05,
+ "loss": 1.224,
+ "step": 582
+ },
+ {
+ "epoch": 3.26,
+ "learning_rate": 3.715083798882682e-05,
+ "loss": 1.3554,
+ "step": 583
+ },
+ {
+ "epoch": 3.26,
+ "learning_rate": 3.687150837988827e-05,
+ "loss": 1.5137,
+ "step": 584
+ },
+ {
+ "epoch": 3.27,
+ "learning_rate": 3.659217877094972e-05,
+ "loss": 1.2462,
+ "step": 585
+ },
+ {
+ "epoch": 3.27,
+ "learning_rate": 3.6312849162011175e-05,
+ "loss": 1.2217,
+ "step": 586
+ },
+ {
+ "epoch": 3.28,
+ "learning_rate": 3.603351955307263e-05,
+ "loss": 1.2346,
+ "step": 587
+ },
+ {
+ "epoch": 3.28,
+ "learning_rate": 3.575418994413408e-05,
+ "loss": 1.3059,
+ "step": 588
+ },
+ {
+ "epoch": 3.29,
+ "learning_rate": 3.5474860335195534e-05,
+ "loss": 1.2355,
+ "step": 589
+ },
+ {
+ "epoch": 3.3,
+ "learning_rate": 3.519553072625699e-05,
+ "loss": 1.2233,
+ "step": 590
+ },
+ {
+ "epoch": 3.3,
+ "learning_rate": 3.491620111731844e-05,
+ "loss": 1.2014,
+ "step": 591
+ },
+ {
+ "epoch": 3.31,
+ "learning_rate": 3.463687150837989e-05,
+ "loss": 1.1919,
+ "step": 592
+ },
+ {
+ "epoch": 3.31,
+ "learning_rate": 3.4357541899441345e-05,
+ "loss": 1.2905,
+ "step": 593
+ },
+ {
+ "epoch": 3.32,
+ "learning_rate": 3.407821229050279e-05,
+ "loss": 1.0858,
+ "step": 594
+ },
+ {
+ "epoch": 3.32,
+ "learning_rate": 3.3798882681564244e-05,
+ "loss": 1.087,
+ "step": 595
+ },
+ {
+ "epoch": 3.33,
+ "learning_rate": 3.35195530726257e-05,
+ "loss": 1.1205,
+ "step": 596
+ },
+ {
+ "epoch": 3.34,
+ "learning_rate": 3.324022346368715e-05,
+ "loss": 1.1657,
+ "step": 597
+ },
+ {
+ "epoch": 3.34,
+ "learning_rate": 3.29608938547486e-05,
+ "loss": 1.5024,
+ "step": 598
+ },
+ {
+ "epoch": 3.35,
+ "learning_rate": 3.2681564245810056e-05,
+ "loss": 1.246,
+ "step": 599
+ },
+ {
+ "epoch": 3.35,
+ "learning_rate": 3.240223463687151e-05,
+ "loss": 1.4441,
+ "step": 600
+ },
+ {
+ "epoch": 3.36,
+ "learning_rate": 3.212290502793296e-05,
+ "loss": 1.19,
+ "step": 601
+ },
+ {
+ "epoch": 3.36,
+ "learning_rate": 3.1843575418994415e-05,
+ "loss": 1.2521,
+ "step": 602
+ },
+ {
+ "epoch": 3.37,
+ "learning_rate": 3.156424581005587e-05,
+ "loss": 1.2051,
+ "step": 603
+ },
+ {
+ "epoch": 3.37,
+ "learning_rate": 3.128491620111732e-05,
+ "loss": 1.4349,
+ "step": 604
+ },
+ {
+ "epoch": 3.38,
+ "learning_rate": 3.1005586592178774e-05,
+ "loss": 0.9633,
+ "step": 605
+ },
+ {
+ "epoch": 3.39,
+ "learning_rate": 3.0726256983240227e-05,
+ "loss": 1.4547,
+ "step": 606
+ },
+ {
+ "epoch": 3.39,
+ "learning_rate": 3.0446927374301676e-05,
+ "loss": 1.2496,
+ "step": 607
+ },
+ {
+ "epoch": 3.4,
+ "learning_rate": 3.0167597765363132e-05,
+ "loss": 1.4193,
+ "step": 608
+ },
+ {
+ "epoch": 3.4,
+ "learning_rate": 2.9888268156424582e-05,
+ "loss": 1.3427,
+ "step": 609
+ },
+ {
+ "epoch": 3.41,
+ "learning_rate": 2.9608938547486038e-05,
+ "loss": 1.3162,
+ "step": 610
+ },
+ {
+ "epoch": 3.41,
+ "learning_rate": 2.9329608938547488e-05,
+ "loss": 1.2194,
+ "step": 611
+ },
+ {
+ "epoch": 3.42,
+ "learning_rate": 2.9050279329608944e-05,
+ "loss": 1.3005,
+ "step": 612
+ },
+ {
+ "epoch": 3.42,
+ "learning_rate": 2.8770949720670394e-05,
+ "loss": 1.2581,
+ "step": 613
+ },
+ {
+ "epoch": 3.43,
+ "learning_rate": 2.8491620111731843e-05,
+ "loss": 1.1465,
+ "step": 614
+ },
+ {
+ "epoch": 3.44,
+ "learning_rate": 2.82122905027933e-05,
+ "loss": 1.1334,
+ "step": 615
+ },
+ {
+ "epoch": 3.44,
+ "learning_rate": 2.793296089385475e-05,
+ "loss": 1.3151,
+ "step": 616
+ },
+ {
+ "epoch": 3.45,
+ "learning_rate": 2.7653631284916205e-05,
+ "loss": 1.3295,
+ "step": 617
+ },
+ {
+ "epoch": 3.45,
+ "learning_rate": 2.7374301675977655e-05,
+ "loss": 1.0037,
+ "step": 618
+ },
+ {
+ "epoch": 3.46,
+ "learning_rate": 2.7094972067039108e-05,
+ "loss": 1.2604,
+ "step": 619
+ },
+ {
+ "epoch": 3.46,
+ "learning_rate": 2.681564245810056e-05,
+ "loss": 1.0984,
+ "step": 620
+ },
+ {
+ "epoch": 3.47,
+ "learning_rate": 2.6536312849162014e-05,
+ "loss": 1.3646,
+ "step": 621
+ },
+ {
+ "epoch": 3.47,
+ "learning_rate": 2.6256983240223466e-05,
+ "loss": 1.3652,
+ "step": 622
+ },
+ {
+ "epoch": 3.48,
+ "learning_rate": 2.5977653631284916e-05,
+ "loss": 1.355,
+ "step": 623
+ },
+ {
+ "epoch": 3.49,
+ "learning_rate": 2.569832402234637e-05,
+ "loss": 1.2907,
+ "step": 624
+ },
+ {
+ "epoch": 3.49,
+ "learning_rate": 2.5418994413407822e-05,
+ "loss": 1.2126,
+ "step": 625
+ },
+ {
+ "epoch": 3.5,
+ "learning_rate": 2.5139664804469275e-05,
+ "loss": 1.1342,
+ "step": 626
+ },
+ {
+ "epoch": 3.5,
+ "learning_rate": 2.4860335195530728e-05,
+ "loss": 1.1031,
+ "step": 627
+ },
+ {
+ "epoch": 3.51,
+ "learning_rate": 2.458100558659218e-05,
+ "loss": 1.2255,
+ "step": 628
+ },
+ {
+ "epoch": 3.51,
+ "learning_rate": 2.4301675977653633e-05,
+ "loss": 1.2866,
+ "step": 629
+ },
+ {
+ "epoch": 3.52,
+ "learning_rate": 2.4022346368715083e-05,
+ "loss": 1.1543,
+ "step": 630
+ },
+ {
+ "epoch": 3.53,
+ "learning_rate": 2.3743016759776536e-05,
+ "loss": 1.2531,
+ "step": 631
+ },
+ {
+ "epoch": 3.53,
+ "learning_rate": 2.346368715083799e-05,
+ "loss": 1.2455,
+ "step": 632
+ },
+ {
+ "epoch": 3.54,
+ "learning_rate": 2.3184357541899442e-05,
+ "loss": 1.2035,
+ "step": 633
+ },
+ {
+ "epoch": 3.54,
+ "learning_rate": 2.2905027932960895e-05,
+ "loss": 1.3482,
+ "step": 634
+ },
+ {
+ "epoch": 3.55,
+ "learning_rate": 2.2625698324022348e-05,
+ "loss": 1.1586,
+ "step": 635
+ },
+ {
+ "epoch": 3.55,
+ "learning_rate": 2.2346368715083797e-05,
+ "loss": 1.1665,
+ "step": 636
+ },
+ {
+ "epoch": 3.56,
+ "learning_rate": 2.206703910614525e-05,
+ "loss": 1.2767,
+ "step": 637
+ },
+ {
+ "epoch": 3.56,
+ "learning_rate": 2.1787709497206703e-05,
+ "loss": 1.302,
+ "step": 638
+ },
+ {
+ "epoch": 3.57,
+ "learning_rate": 2.1508379888268156e-05,
+ "loss": 1.2224,
+ "step": 639
+ },
+ {
+ "epoch": 3.58,
+ "learning_rate": 2.122905027932961e-05,
+ "loss": 1.1817,
+ "step": 640
+ },
+ {
+ "epoch": 3.58,
+ "learning_rate": 2.0949720670391062e-05,
+ "loss": 1.4949,
+ "step": 641
+ },
+ {
+ "epoch": 3.59,
+ "learning_rate": 2.0670391061452515e-05,
+ "loss": 1.2552,
+ "step": 642
+ },
+ {
+ "epoch": 3.59,
+ "learning_rate": 2.0391061452513968e-05,
+ "loss": 1.3075,
+ "step": 643
+ },
+ {
+ "epoch": 3.6,
+ "learning_rate": 2.011173184357542e-05,
+ "loss": 1.1825,
+ "step": 644
+ },
+ {
+ "epoch": 3.6,
+ "learning_rate": 1.9832402234636873e-05,
+ "loss": 1.484,
+ "step": 645
+ },
+ {
+ "epoch": 3.61,
+ "learning_rate": 1.9553072625698323e-05,
+ "loss": 1.1765,
+ "step": 646
+ },
+ {
+ "epoch": 3.61,
+ "learning_rate": 1.9273743016759776e-05,
+ "loss": 1.381,
+ "step": 647
+ },
+ {
+ "epoch": 3.62,
+ "learning_rate": 1.899441340782123e-05,
+ "loss": 1.2835,
+ "step": 648
+ },
+ {
+ "epoch": 3.63,
+ "learning_rate": 1.871508379888268e-05,
+ "loss": 1.235,
+ "step": 649
+ },
+ {
+ "epoch": 3.63,
+ "learning_rate": 1.8435754189944135e-05,
+ "loss": 1.2041,
+ "step": 650
+ },
+ {
+ "epoch": 3.64,
+ "learning_rate": 1.8156424581005588e-05,
+ "loss": 0.9949,
+ "step": 651
+ },
+ {
+ "epoch": 3.64,
+ "learning_rate": 1.787709497206704e-05,
+ "loss": 1.2147,
+ "step": 652
+ },
+ {
+ "epoch": 3.65,
+ "learning_rate": 1.7597765363128493e-05,
+ "loss": 1.3876,
+ "step": 653
+ },
+ {
+ "epoch": 3.65,
+ "learning_rate": 1.7318435754189946e-05,
+ "loss": 1.1255,
+ "step": 654
+ },
+ {
+ "epoch": 3.66,
+ "learning_rate": 1.7039106145251396e-05,
+ "loss": 1.2252,
+ "step": 655
+ },
+ {
+ "epoch": 3.66,
+ "learning_rate": 1.675977653631285e-05,
+ "loss": 1.303,
+ "step": 656
+ },
+ {
+ "epoch": 3.67,
+ "learning_rate": 1.64804469273743e-05,
+ "loss": 1.1023,
+ "step": 657
+ },
+ {
+ "epoch": 3.68,
+ "learning_rate": 1.6201117318435755e-05,
+ "loss": 1.1444,
+ "step": 658
+ },
+ {
+ "epoch": 3.68,
+ "learning_rate": 1.5921787709497207e-05,
+ "loss": 1.1893,
+ "step": 659
+ },
+ {
+ "epoch": 3.69,
+ "learning_rate": 1.564245810055866e-05,
+ "loss": 1.2557,
+ "step": 660
+ },
+ {
+ "epoch": 3.69,
+ "learning_rate": 1.5363128491620113e-05,
+ "loss": 1.3431,
+ "step": 661
+ },
+ {
+ "epoch": 3.7,
+ "learning_rate": 1.5083798882681566e-05,
+ "loss": 1.389,
+ "step": 662
+ },
+ {
+ "epoch": 3.7,
+ "learning_rate": 1.4804469273743019e-05,
+ "loss": 1.0841,
+ "step": 663
+ },
+ {
+ "epoch": 3.71,
+ "learning_rate": 1.4525139664804472e-05,
+ "loss": 1.1225,
+ "step": 664
+ },
+ {
+ "epoch": 3.72,
+ "learning_rate": 1.4245810055865922e-05,
+ "loss": 1.2914,
+ "step": 665
+ },
+ {
+ "epoch": 3.72,
+ "learning_rate": 1.3966480446927374e-05,
+ "loss": 1.2876,
+ "step": 666
+ },
+ {
+ "epoch": 3.73,
+ "learning_rate": 1.3687150837988827e-05,
+ "loss": 1.2285,
+ "step": 667
+ },
+ {
+ "epoch": 3.73,
+ "learning_rate": 1.340782122905028e-05,
+ "loss": 1.0809,
+ "step": 668
+ },
+ {
+ "epoch": 3.74,
+ "learning_rate": 1.3128491620111733e-05,
+ "loss": 1.0506,
+ "step": 669
+ },
+ {
+ "epoch": 3.74,
+ "learning_rate": 1.2849162011173184e-05,
+ "loss": 1.2041,
+ "step": 670
+ },
+ {
+ "epoch": 3.75,
+ "learning_rate": 1.2569832402234637e-05,
+ "loss": 1.1061,
+ "step": 671
+ },
+ {
+ "epoch": 3.75,
+ "learning_rate": 1.229050279329609e-05,
+ "loss": 1.2219,
+ "step": 672
+ },
+ {
+ "epoch": 3.76,
+ "learning_rate": 1.2011173184357542e-05,
+ "loss": 1.0946,
+ "step": 673
+ },
+ {
+ "epoch": 3.77,
+ "learning_rate": 1.1731843575418994e-05,
+ "loss": 1.089,
+ "step": 674
+ },
+ {
+ "epoch": 3.77,
+ "learning_rate": 1.1452513966480447e-05,
+ "loss": 1.237,
+ "step": 675
+ },
+ {
+ "epoch": 3.78,
+ "learning_rate": 1.1173184357541899e-05,
+ "loss": 1.4085,
+ "step": 676
+ },
+ {
+ "epoch": 3.78,
+ "learning_rate": 1.0893854748603351e-05,
+ "loss": 1.2196,
+ "step": 677
+ },
+ {
+ "epoch": 3.79,
+ "learning_rate": 1.0614525139664804e-05,
+ "loss": 1.1506,
+ "step": 678
+ },
+ {
+ "epoch": 3.79,
+ "learning_rate": 1.0335195530726257e-05,
+ "loss": 1.2791,
+ "step": 679
+ },
+ {
+ "epoch": 3.8,
+ "learning_rate": 1.005586592178771e-05,
+ "loss": 1.3344,
+ "step": 680
+ },
+ {
+ "epoch": 3.8,
+ "learning_rate": 9.776536312849161e-06,
+ "loss": 1.3458,
+ "step": 681
+ },
+ {
+ "epoch": 3.81,
+ "learning_rate": 9.497206703910614e-06,
+ "loss": 1.1622,
+ "step": 682
+ },
+ {
+ "epoch": 3.82,
+ "learning_rate": 9.217877094972067e-06,
+ "loss": 1.1179,
+ "step": 683
+ },
+ {
+ "epoch": 3.82,
+ "learning_rate": 8.93854748603352e-06,
+ "loss": 1.2485,
+ "step": 684
+ },
+ {
+ "epoch": 3.83,
+ "learning_rate": 8.659217877094973e-06,
+ "loss": 1.1207,
+ "step": 685
+ },
+ {
+ "epoch": 3.83,
+ "learning_rate": 8.379888268156424e-06,
+ "loss": 1.3791,
+ "step": 686
+ },
+ {
+ "epoch": 3.84,
+ "learning_rate": 8.100558659217877e-06,
+ "loss": 1.2976,
+ "step": 687
+ },
+ {
+ "epoch": 3.84,
+ "learning_rate": 7.82122905027933e-06,
+ "loss": 1.2041,
+ "step": 688
+ },
+ {
+ "epoch": 3.85,
+ "learning_rate": 7.541899441340783e-06,
+ "loss": 1.1002,
+ "step": 689
+ },
+ {
+ "epoch": 3.85,
+ "learning_rate": 7.262569832402236e-06,
+ "loss": 1.2982,
+ "step": 690
+ },
+ {
+ "epoch": 3.86,
+ "learning_rate": 6.983240223463687e-06,
+ "loss": 1.3469,
+ "step": 691
+ },
+ {
+ "epoch": 3.87,
+ "learning_rate": 6.70391061452514e-06,
+ "loss": 1.2923,
+ "step": 692
+ },
+ {
+ "epoch": 3.87,
+ "learning_rate": 6.424581005586592e-06,
+ "loss": 1.3308,
+ "step": 693
+ },
+ {
+ "epoch": 3.88,
+ "learning_rate": 6.145251396648045e-06,
+ "loss": 1.2588,
+ "step": 694
+ },
+ {
+ "epoch": 3.88,
+ "learning_rate": 5.865921787709497e-06,
+ "loss": 1.0669,
+ "step": 695
+ },
+ {
+ "epoch": 3.89,
+ "learning_rate": 5.586592178770949e-06,
+ "loss": 1.4808,
+ "step": 696
+ },
+ {
+ "epoch": 3.89,
+ "learning_rate": 5.307262569832402e-06,
+ "loss": 1.2714,
+ "step": 697
+ },
+ {
+ "epoch": 3.9,
+ "learning_rate": 5.027932960893855e-06,
+ "loss": 1.1979,
+ "step": 698
+ },
+ {
+ "epoch": 3.91,
+ "learning_rate": 4.748603351955307e-06,
+ "loss": 1.2548,
+ "step": 699
+ },
+ {
+ "epoch": 3.91,
+ "learning_rate": 4.46927374301676e-06,
+ "loss": 1.2473,
+ "step": 700
+ }
+ ],
+ "logging_steps": 1,
+ "max_steps": 716,
+ "num_train_epochs": 4,
+ "save_steps": 100,
+ "total_flos": 3.6028388613095424e+17,
+ "trial_name": null,
+ "trial_params": null
+}
diff --git a/checkpoint-700/training_args.bin b/checkpoint-700/training_args.bin
new file mode 100644
index 0000000000000000000000000000000000000000..c8672c716e925d0028b4938db147703f58656ff7
--- /dev/null
+++ b/checkpoint-700/training_args.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:188ae1c421cc0c6435d1f71d8d3423ac4abc7dba0e6fc2efcbc4dbe77c741317
+size 4027