Training in progress, epoch 9, checkpoint

checkpoint-2250/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.5.0

checkpoint-2250/adapter_config.json

@@ -0,0 +1,25 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "Qwen/Qwen1.5-4B",
+  "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": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "k_proj",
+    "v_proj",
+    "down_proj",
+    "gate_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-2250/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:affadb8b31179b0cd863673a15cfbd309333faa6684aecc725ea0eb4bedf2a98
+size 112130416

checkpoint-2250/added_tokens.json

@@ -0,0 +1,5 @@
+{
+  "<|endoftext|>": 151643,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644
+}

checkpoint-2250/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6402b11d541cd02ccf6936e6b59b7f71b1aab9d8fcd8797106d22fbcf2ded90a
+size 224537202

checkpoint-2250/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10ad9d3c8da575eab4dce1d0fd27faae8881ed2f8b7d824ea91554972be14202
+size 15024

checkpoint-2250/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc2e1e7c7596fa8ac9a427794b880699469ff72097fa6a7045634b6b2a1944eb
+size 15024

checkpoint-2250/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b5c3866d0d8c1601c3bbe0f09e3030991510b4b05cd9d8334e676828f55455e
+size 15024

checkpoint-2250/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcbaf3716a263b8a7d4f7d20c10c643931b00647e2ea4628fb6fa5b762283d73
+size 15024

checkpoint-2250/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c204508ddabfa22d7f1896a9cb8b8476930bb6588260caab49291a4b903cc826
+size 1064

checkpoint-2250/special_tokens_map.json

@@ -0,0 +1,14 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "eos_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<|endoftext|>"
+}

checkpoint-2250/tokenizer_config.json

@@ -0,0 +1,43 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

checkpoint-2250/trainer_state.json

@@ -0,0 +1,322 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.0,
+  "eval_steps": 500,
+  "global_step": 2250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5479037761688232,
+      "learning_rate": 0.0005,
+      "loss": 2.3029,
+      "step": 100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5442906618118286,
+      "learning_rate": 0.0005,
+      "loss": 2.2398,
+      "step": 200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5162539682539683,
+      "eval_loss": 2.3236427307128906,
+      "eval_runtime": 5.266,
+      "eval_samples_per_second": 94.949,
+      "eval_steps_per_second": 11.964,
+      "step": 250
+    },
+    {
+      "epoch": 1.0,
+      "eval_exact_match": 17.6,
+      "eval_f1": 26.59450937950939,
+      "step": 250
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6918729543685913,
+      "learning_rate": 0.0005,
+      "loss": 2.0245,
+      "step": 300
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7950774431228638,
+      "learning_rate": 0.0005,
+      "loss": 1.8286,
+      "step": 400
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6913579106330872,
+      "learning_rate": 0.0005,
+      "loss": 1.8301,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5124126984126984,
+      "eval_loss": 2.4219954013824463,
+      "eval_runtime": 5.2763,
+      "eval_samples_per_second": 94.763,
+      "eval_steps_per_second": 11.94,
+      "step": 500
+    },
+    {
+      "epoch": 2.0,
+      "eval_exact_match": 17.0,
+      "eval_f1": 24.808081363081367,
+      "step": 500
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.9923084378242493,
+      "learning_rate": 0.0005,
+      "loss": 1.3144,
+      "step": 600
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.0025588274002075,
+      "learning_rate": 0.0005,
+      "loss": 1.3626,
+      "step": 700
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5061904761904762,
+      "eval_loss": 2.6152632236480713,
+      "eval_runtime": 5.2745,
+      "eval_samples_per_second": 94.795,
+      "eval_steps_per_second": 11.944,
+      "step": 750
+    },
+    {
+      "epoch": 3.0,
+      "eval_exact_match": 16.0,
+      "eval_f1": 24.234217449217454,
+      "step": 750
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.0352345705032349,
+      "learning_rate": 0.0005,
+      "loss": 1.1681,
+      "step": 800
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.1078394651412964,
+      "learning_rate": 0.0005,
+      "loss": 0.9584,
+      "step": 900
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.14862060546875,
+      "learning_rate": 0.0005,
+      "loss": 1.0112,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.49974603174603177,
+      "eval_loss": 2.834880828857422,
+      "eval_runtime": 5.1472,
+      "eval_samples_per_second": 97.141,
+      "eval_steps_per_second": 12.24,
+      "step": 1000
+    },
+    {
+      "epoch": 4.0,
+      "eval_exact_match": 13.6,
+      "eval_f1": 20.6380957930958,
+      "step": 1000
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.2231959104537964,
+      "learning_rate": 0.0005,
+      "loss": 0.6614,
+      "step": 1100
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.1669212579727173,
+      "learning_rate": 0.0005,
+      "loss": 0.7198,
+      "step": 1200
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.4963174603174603,
+      "eval_loss": 3.075563907623291,
+      "eval_runtime": 5.1584,
+      "eval_samples_per_second": 96.929,
+      "eval_steps_per_second": 12.213,
+      "step": 1250
+    },
+    {
+      "epoch": 5.0,
+      "eval_exact_match": 14.0,
+      "eval_f1": 20.88632090132091,
+      "step": 1250
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.9650686979293823,
+      "learning_rate": 0.0005,
+      "loss": 0.6299,
+      "step": 1300
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.9634979963302612,
+      "learning_rate": 0.0005,
+      "loss": 0.5491,
+      "step": 1400
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.0452007055282593,
+      "learning_rate": 0.0005,
+      "loss": 0.589,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.49434920634920637,
+      "eval_loss": 3.2339136600494385,
+      "eval_runtime": 5.1512,
+      "eval_samples_per_second": 97.064,
+      "eval_steps_per_second": 12.23,
+      "step": 1500
+    },
+    {
+      "epoch": 6.0,
+      "eval_exact_match": 15.0,
+      "eval_f1": 21.66983199806731,
+      "step": 1500
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 1.0834842920303345,
+      "learning_rate": 0.0005,
+      "loss": 0.4679,
+      "step": 1600
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.9649925231933594,
+      "learning_rate": 0.0005,
+      "loss": 0.4969,
+      "step": 1700
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.4934920634920635,
+      "eval_loss": 3.34251070022583,
+      "eval_runtime": 5.2951,
+      "eval_samples_per_second": 94.426,
+      "eval_steps_per_second": 11.898,
+      "step": 1750
+    },
+    {
+      "epoch": 7.0,
+      "eval_exact_match": 13.8,
+      "eval_f1": 20.9043167943168,
+      "step": 1750
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.8583429455757141,
+      "learning_rate": 0.0005,
+      "loss": 0.4737,
+      "step": 1800
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.9595094919204712,
+      "learning_rate": 0.0005,
+      "loss": 0.4544,
+      "step": 1900
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.9260773062705994,
+      "learning_rate": 0.0005,
+      "loss": 0.4786,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.49244444444444446,
+      "eval_loss": 3.4198057651519775,
+      "eval_runtime": 5.8813,
+      "eval_samples_per_second": 85.015,
+      "eval_steps_per_second": 10.712,
+      "step": 2000
+    },
+    {
+      "epoch": 8.0,
+      "eval_exact_match": 13.4,
+      "eval_f1": 20.923260628260632,
+      "step": 2000
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 1.045862078666687,
+      "learning_rate": 0.0005,
+      "loss": 0.4114,
+      "step": 2100
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.7850492000579834,
+      "learning_rate": 0.0005,
+      "loss": 0.4399,
+      "step": 2200
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.49114285714285716,
+      "eval_loss": 3.4694714546203613,
+      "eval_runtime": 5.945,
+      "eval_samples_per_second": 84.104,
+      "eval_steps_per_second": 10.597,
+      "step": 2250
+    },
+    {
+      "epoch": 9.0,
+      "eval_exact_match": 13.0,
+      "eval_f1": 20.096701018279973,
+      "step": 2250
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 12500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.881574399875482e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-2250/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b62f42960cf4fd93abfb3be317e2e8ef04964ff17a18b3862335bf3c3ff7041
+size 5304