Training in progress, step 200

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",
+ "gate_proj",
+ "up_proj",
+ "down_proj",
+ "k_proj",
+ "q_proj",
+ "v_proj"
+ ],
+ "task_type": "CAUSAL_LM"
+}

adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4bcd2b46f70d5681370d5cb5eaa04a971592bc85bb67124e9c3a643fdbdef48
+size 319977229

added_tokens.json

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

checkpoint-200/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-200/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",
+ "gate_proj",
+ "up_proj",
+ "down_proj",
+ "k_proj",
+ "q_proj",
+ "v_proj"
+ ],
+ "task_type": "CAUSAL_LM"
+}

checkpoint-200/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4bcd2b46f70d5681370d5cb5eaa04a971592bc85bb67124e9c3a643fdbdef48
+size 319977229

checkpoint-200/added_tokens.json

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

checkpoint-200/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a650394b7dfe157dff041d90839cbb47a386dfb8bb7ef00aebe2cb05d07abfce
+size 1279539525

checkpoint-200/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f078a5ef09c426d3a6069669845a570c0c9ce0e30b14901295b4d1005c6c2e7a
+size 14511

checkpoint-200/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7aa2c8b84e17817e6a4dcba5955fca913e266fdcd47f5594a29933ebd4972a01
+size 627

checkpoint-200/special_tokens_map.json

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

checkpoint-200/tokenizer.model

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

checkpoint-200/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-200/trainer_state.json

@@ -0,0 +1,207 @@
+{
+ "best_metric": 0.6798371076583862,
+ "best_model_checkpoint": "./output_v2/7b_cluster012_Nous-Hermes-llama-2-7b_partitioned_v3_standardized_012/checkpoint-200",
+ "epoch": 0.23161551823972207,
+ "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": 0.0002,
+ "loss": 0.7418,
+ "step": 10
+ },
+ {
+ "epoch": 0.02,
+ "learning_rate": 0.0002,
+ "loss": 0.7269,
+ "step": 20
+ },
+ {
+ "epoch": 0.03,
+ "learning_rate": 0.0002,
+ "loss": 0.7061,
+ "step": 30
+ },
+ {
+ "epoch": 0.05,
+ "learning_rate": 0.0002,
+ "loss": 0.6809,
+ "step": 40
+ },
+ {
+ "epoch": 0.06,
+ "learning_rate": 0.0002,
+ "loss": 0.6742,
+ "step": 50
+ },
+ {
+ "epoch": 0.07,
+ "learning_rate": 0.0002,
+ "loss": 0.6789,
+ "step": 60
+ },
+ {
+ "epoch": 0.08,
+ "learning_rate": 0.0002,
+ "loss": 0.6628,
+ "step": 70
+ },
+ {
+ "epoch": 0.09,
+ "learning_rate": 0.0002,
+ "loss": 0.6995,
+ "step": 80
+ },
+ {
+ "epoch": 0.1,
+ "learning_rate": 0.0002,
+ "loss": 0.6774,
+ "step": 90
+ },
+ {
+ "epoch": 0.12,
+ "learning_rate": 0.0002,
+ "loss": 0.6716,
+ "step": 100
+ },
+ {
+ "epoch": 0.13,
+ "learning_rate": 0.0002,
+ "loss": 0.6628,
+ "step": 110
+ },
+ {
+ "epoch": 0.14,
+ "learning_rate": 0.0002,
+ "loss": 0.6696,
+ "step": 120
+ },
+ {
+ "epoch": 0.15,
+ "learning_rate": 0.0002,
+ "loss": 0.648,
+ "step": 130
+ },
+ {
+ "epoch": 0.16,
+ "learning_rate": 0.0002,
+ "loss": 0.6823,
+ "step": 140
+ },
+ {
+ "epoch": 0.17,
+ "learning_rate": 0.0002,
+ "loss": 0.6532,
+ "step": 150
+ },
+ {
+ "epoch": 0.19,
+ "learning_rate": 0.0002,
+ "loss": 0.669,
+ "step": 160
+ },
+ {
+ "epoch": 0.2,
+ "learning_rate": 0.0002,
+ "loss": 0.6639,
+ "step": 170
+ },
+ {
+ "epoch": 0.21,
+ "learning_rate": 0.0002,
+ "loss": 0.6582,
+ "step": 180
+ },
+ {
+ "epoch": 0.22,
+ "learning_rate": 0.0002,
+ "loss": 0.6381,
+ "step": 190
+ },
+ {
+ "epoch": 0.23,
+ "learning_rate": 0.0002,
+ "loss": 0.6342,
+ "step": 200
+ },
+ {
+ "epoch": 0.23,
+ "eval_loss": 0.6798371076583862,
+ "eval_runtime": 220.8775,
+ "eval_samples_per_second": 4.527,
+ "eval_steps_per_second": 2.264,
+ "step": 200
+ },
+ {
+ "epoch": 0.23,
+ "mmlu_eval_accuracy": 0.4661323913730994,
+ "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": 0.4375,
+ "mmlu_eval_accuracy_college_chemistry": 0.125,
+ "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
+ "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
+ "mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
+ "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": 0.3125,
+ "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
+ "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
+ "mmlu_eval_accuracy_global_facts": 0.5,
+ "mmlu_eval_accuracy_high_school_biology": 0.375,
+ "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
+ "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
+ "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
+ "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
+ "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
+ "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
+ "mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
+ "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
+ "mmlu_eval_accuracy_high_school_psychology": 0.7166666666666667,
+ "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
+ "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
+ "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
+ "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": 0.2727272727272727,
+ "mmlu_eval_accuracy_management": 0.6363636363636364,
+ "mmlu_eval_accuracy_marketing": 0.76,
+ "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
+ "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
+ "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
+ "mmlu_eval_accuracy_moral_scenarios": 0.24,
+ "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
+ "mmlu_eval_accuracy_philosophy": 0.5,
+ "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
+ "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
+ "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
+ "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
+ "mmlu_eval_accuracy_professional_psychology": 0.391304347826087,
+ "mmlu_eval_accuracy_public_relations": 0.5,
+ "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
+ "mmlu_eval_accuracy_sociology": 0.6818181818181818,
+ "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
+ "mmlu_eval_accuracy_virology": 0.3888888888888889,
+ "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
+ "mmlu_loss": 1.1315238676556718,
+ "step": 200
+ }
+ ],
+ "max_steps": 5000,
+ "num_train_epochs": 6,
+ "total_flos": 5.082084345131827e+16,
+ "trial_name": null,
+ "trial_params": null
+}

checkpoint-200/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de1718044df058d493a02df580cf2d03a18301bd3fb5c130230ece0994222cf6
+size 6011

special_tokens_map.json

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

tokenizer.model

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

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
+ }
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de1718044df058d493a02df580cf2d03a18301bd3fb5c130230ece0994222cf6
+size 6011