Training in progress, step 600

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea25c1bdefaabf0a8e23d433facecb6ce5bbc1bc5a718607a92a90b6daafb18a
+oid sha256:a48621448d0af006faeb8c5389a587f73f9b6385da07316290db9c169e7beb25
 size 319977229

checkpoint-400/adapter_model/adapter_model/README.md

@@ -0,0 +1,20 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- load_in_8bit: False
+- load_in_4bit: True
+- 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: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: bfloat16
+### Framework versions
+
+
+- PEFT 0.4.0

checkpoint-400/adapter_model/adapter_model/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "NousResearch/Nous-Hermes-llama-2-7b",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16.0,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "q_proj",
+    "down_proj",
+    "up_proj",
+    "gate_proj",
+    "v_proj",
+    "k_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-400/adapter_model/adapter_model/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea25c1bdefaabf0a8e23d433facecb6ce5bbc1bc5a718607a92a90b6daafb18a
+size 319977229

checkpoint-600/README.md

@@ -0,0 +1,20 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- load_in_8bit: False
+- load_in_4bit: True
+- 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: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: bfloat16
+### Framework versions
+
+
+- PEFT 0.4.0

checkpoint-600/adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "NousResearch/Nous-Hermes-llama-2-7b",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16.0,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "q_proj",
+    "down_proj",
+    "up_proj",
+    "gate_proj",
+    "v_proj",
+    "k_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-600/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a48621448d0af006faeb8c5389a587f73f9b6385da07316290db9c169e7beb25
+size 319977229

checkpoint-600/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<pad>": 32000
+}

checkpoint-600/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b704402ea4c797395d903ce5fbc5099558a66938c5dede20ea5a17445075a606
+size 1279539973

checkpoint-600/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae8ec1441bcb7cbcb267d6fe424477aa507af656f1c130627ca7fd9b6bb1adaf
+size 14511

checkpoint-600/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c1a2ac1c11599601eeac95feb1dbfd49ec5c625e61dcce18b3f094491f9cf2d
+size 627

checkpoint-600/special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "<unk>",
+  "unk_token": "<unk>"
+}

checkpoint-600/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

checkpoint-600/tokenizer_config.json

@@ -0,0 +1,35 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "legacy": false,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": null,
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-600/trainer_state.json

@@ -0,0 +1,589 @@
+{
+  "best_metric": 1.0200624465942383,
+  "best_model_checkpoint": "./output_v2/7b_cluster07_Nous-Hermes-llama-2-7b_partitioned_v3_standardized_07/checkpoint-600",
+  "epoch": 0.3815580286168522,
+  "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": 0.0002,
+      "loss": 1.0736,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002,
+      "loss": 1.1041,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002,
+      "loss": 1.0818,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002,
+      "loss": 1.0408,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002,
+      "loss": 1.0985,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002,
+      "loss": 1.0245,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002,
+      "loss": 1.0205,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002,
+      "loss": 1.0811,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002,
+      "loss": 1.0852,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002,
+      "loss": 1.0296,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002,
+      "loss": 1.0943,
+      "step": 110
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002,
+      "loss": 0.9857,
+      "step": 120
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002,
+      "loss": 1.0324,
+      "step": 130
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002,
+      "loss": 1.0134,
+      "step": 140
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002,
+      "loss": 1.0533,
+      "step": 150
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002,
+      "loss": 1.0667,
+      "step": 160
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002,
+      "loss": 1.0506,
+      "step": 170
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002,
+      "loss": 1.0653,
+      "step": 180
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002,
+      "loss": 1.0372,
+      "step": 190
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002,
+      "loss": 1.0485,
+      "step": 200
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 1.0341941118240356,
+      "eval_runtime": 172.5264,
+      "eval_samples_per_second": 5.796,
+      "eval_steps_per_second": 2.898,
+      "step": 200
+    },
+    {
+      "epoch": 0.13,
+      "mmlu_eval_accuracy": 0.4648810025502313,
+      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
+      "mmlu_eval_accuracy_anatomy": 0.5,
+      "mmlu_eval_accuracy_astronomy": 0.375,
+      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
+      "mmlu_eval_accuracy_college_biology": 0.375,
+      "mmlu_eval_accuracy_college_chemistry": 0.125,
+      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
+      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
+      "mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
+      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
+      "mmlu_eval_accuracy_computer_security": 0.36363636363636365,
+      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
+      "mmlu_eval_accuracy_econometrics": 0.25,
+      "mmlu_eval_accuracy_electrical_engineering": 0.5,
+      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
+      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
+      "mmlu_eval_accuracy_global_facts": 0.5,
+      "mmlu_eval_accuracy_high_school_biology": 0.34375,
+      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
+      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
+      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
+      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
+      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
+      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
+      "mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
+      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+      "mmlu_eval_accuracy_high_school_psychology": 0.7166666666666667,
+      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
+      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+      "mmlu_eval_accuracy_high_school_world_history": 0.5,
+      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
+      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
+      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
+      "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
+      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
+      "mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
+      "mmlu_eval_accuracy_management": 0.6363636363636364,
+      "mmlu_eval_accuracy_marketing": 0.8,
+      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
+      "mmlu_eval_accuracy_moral_disputes": 0.5,
+      "mmlu_eval_accuracy_moral_scenarios": 0.24,
+      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
+      "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
+      "mmlu_eval_accuracy_prehistory": 0.42857142857142855,
+      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
+      "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
+      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
+      "mmlu_eval_accuracy_professional_psychology": 0.36231884057971014,
+      "mmlu_eval_accuracy_public_relations": 0.5,
+      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
+      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
+      "mmlu_eval_accuracy_us_foreign_policy": 0.45454545454545453,
+      "mmlu_eval_accuracy_virology": 0.4444444444444444,
+      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
+      "mmlu_loss": 1.115347789702621,
+      "step": 200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002,
+      "loss": 0.987,
+      "step": 210
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002,
+      "loss": 1.0399,
+      "step": 220
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002,
+      "loss": 1.044,
+      "step": 230
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002,
+      "loss": 1.0491,
+      "step": 240
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002,
+      "loss": 1.0216,
+      "step": 250
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002,
+      "loss": 1.0973,
+      "step": 260
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002,
+      "loss": 0.996,
+      "step": 270
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002,
+      "loss": 1.0253,
+      "step": 280
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002,
+      "loss": 1.0439,
+      "step": 290
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002,
+      "loss": 1.0244,
+      "step": 300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002,
+      "loss": 1.0299,
+      "step": 310
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002,
+      "loss": 1.0737,
+      "step": 320
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002,
+      "loss": 0.9939,
+      "step": 330
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002,
+      "loss": 1.032,
+      "step": 340
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002,
+      "loss": 1.0291,
+      "step": 350
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002,
+      "loss": 1.0575,
+      "step": 360
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002,
+      "loss": 1.0685,
+      "step": 370
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002,
+      "loss": 1.0342,
+      "step": 380
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 1.0055,
+      "step": 390
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 1.0584,
+      "step": 400
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.0254805088043213,
+      "eval_runtime": 172.7517,
+      "eval_samples_per_second": 5.789,
+      "eval_steps_per_second": 2.894,
+      "step": 400
+    },
+    {
+      "epoch": 0.25,
+      "mmlu_eval_accuracy": 0.4742191978590581,
+      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
+      "mmlu_eval_accuracy_anatomy": 0.5,
+      "mmlu_eval_accuracy_astronomy": 0.5,
+      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
+      "mmlu_eval_accuracy_college_biology": 0.375,
+      "mmlu_eval_accuracy_college_chemistry": 0.125,
+      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
+      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
+      "mmlu_eval_accuracy_college_medicine": 0.2727272727272727,
+      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+      "mmlu_eval_accuracy_computer_security": 0.2727272727272727,
+      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
+      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
+      "mmlu_eval_accuracy_electrical_engineering": 0.5,
+      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
+      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
+      "mmlu_eval_accuracy_global_facts": 0.5,
+      "mmlu_eval_accuracy_high_school_biology": 0.34375,
+      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
+      "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778,
+      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
+      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
+      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
+      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
+      "mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
+      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+      "mmlu_eval_accuracy_high_school_psychology": 0.6833333333333333,
+      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
+      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
+      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+      "mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
+      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
+      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
+      "mmlu_eval_accuracy_logical_fallacies": 0.5,
+      "mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
+      "mmlu_eval_accuracy_management": 0.7272727272727273,
+      "mmlu_eval_accuracy_marketing": 0.72,
+      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
+      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
+      "mmlu_eval_accuracy_moral_scenarios": 0.24,
+      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
+      "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
+      "mmlu_eval_accuracy_prehistory": 0.37142857142857144,
+      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
+      "mmlu_eval_accuracy_professional_law": 0.32941176470588235,
+      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
+      "mmlu_eval_accuracy_professional_psychology": 0.391304347826087,
+      "mmlu_eval_accuracy_public_relations": 0.5,
+      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
+      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
+      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
+      "mmlu_eval_accuracy_virology": 0.6111111111111112,
+      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
+      "mmlu_loss": 1.038680466208072,
+      "step": 400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002,
+      "loss": 1.0107,
+      "step": 410
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002,
+      "loss": 1.0667,
+      "step": 420
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002,
+      "loss": 0.9837,
+      "step": 430
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002,
+      "loss": 1.0534,
+      "step": 440
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002,
+      "loss": 0.9922,
+      "step": 450
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002,
+      "loss": 1.0146,
+      "step": 460
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002,
+      "loss": 1.0438,
+      "step": 470
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002,
+      "loss": 0.9886,
+      "step": 480
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002,
+      "loss": 0.988,
+      "step": 490
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002,
+      "loss": 1.0228,
+      "step": 500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002,
+      "loss": 1.0173,
+      "step": 510
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002,
+      "loss": 0.9993,
+      "step": 520
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002,
+      "loss": 1.0261,
+      "step": 530
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002,
+      "loss": 0.9884,
+      "step": 540
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002,
+      "loss": 0.9894,
+      "step": 550
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002,
+      "loss": 1.0305,
+      "step": 560
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002,
+      "loss": 0.9754,
+      "step": 570
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002,
+      "loss": 1.0075,
+      "step": 580
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002,
+      "loss": 1.0219,
+      "step": 590
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002,
+      "loss": 1.0059,
+      "step": 600
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.0200624465942383,
+      "eval_runtime": 172.8545,
+      "eval_samples_per_second": 5.785,
+      "eval_steps_per_second": 2.893,
+      "step": 600
+    },
+    {
+      "epoch": 0.38,
+      "mmlu_eval_accuracy": 0.46940456315845464,
+      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
+      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
+      "mmlu_eval_accuracy_astronomy": 0.5,
+      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
+      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
+      "mmlu_eval_accuracy_college_biology": 0.375,
+      "mmlu_eval_accuracy_college_chemistry": 0.125,
+      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
+      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
+      "mmlu_eval_accuracy_college_medicine": 0.3181818181818182,
+      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
+      "mmlu_eval_accuracy_computer_security": 0.36363636363636365,
+      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
+      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
+      "mmlu_eval_accuracy_electrical_engineering": 0.5,
+      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
+      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
+      "mmlu_eval_accuracy_global_facts": 0.4,
+      "mmlu_eval_accuracy_high_school_biology": 0.34375,
+      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
+      "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778,
+      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
+      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
+      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
+      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
+      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
+      "mmlu_eval_accuracy_high_school_microeconomics": 0.34615384615384615,
+      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
+      "mmlu_eval_accuracy_high_school_psychology": 0.6833333333333333,
+      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
+      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
+      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
+      "mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
+      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
+      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
+      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
+      "mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
+      "mmlu_eval_accuracy_management": 0.7272727272727273,
+      "mmlu_eval_accuracy_marketing": 0.76,
+      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
+      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
+      "mmlu_eval_accuracy_moral_scenarios": 0.24,
+      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
+      "mmlu_eval_accuracy_philosophy": 0.5,
+      "mmlu_eval_accuracy_prehistory": 0.37142857142857144,
+      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
+      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
+      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
+      "mmlu_eval_accuracy_professional_psychology": 0.42028985507246375,
+      "mmlu_eval_accuracy_public_relations": 0.5,
+      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
+      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
+      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
+      "mmlu_eval_accuracy_virology": 0.4444444444444444,
+      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
+      "mmlu_loss": 1.0641063005121196,
+      "step": 600
+    }
+  ],
+  "max_steps": 5000,
+  "num_train_epochs": 4,
+  "total_flos": 1.1428667466448896e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-600/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfe6da60c477ef786e72ea5b4cb696c083fdb494b274f319fe5be152de00d6ec
+size 6011