Model save

README.md

@@ -0,0 +1,56 @@
+---
+base_model: hllj/mistral-vi-math
+tags:
+- generated_from_trainer
+model-index:
+- name: sft-mistral-v3-all
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# sft-mistral-v3-all
+
+This model is a fine-tuned version of [hllj/mistral-vi-math](https://huggingface.co/hllj/mistral-vi-math) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.4937
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 5e-05
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.05
+- num_epochs: 2
+- mixed_precision_training: Native AMP
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.35.2
+- Pytorch 2.1.0
+- Datasets 2.15.0
+- Tokenizers 0.15.0

adapter_config.json

@@ -0,0 +1,25 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "hllj/mistral-vi-math",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 128,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 256,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "k_proj",
+    "v_proj",
+    "q_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f6bca230ef43184c05ba5106e1b5978b26c04de78285775fc6de35254c0b91e
+size 872450448

all_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 1.45,
+    "eval_loss": 0.4937315583229065,
+    "eval_runtime": 81.9976,
+    "eval_samples": 852,
+    "eval_samples_per_second": 10.391,
+    "eval_steps_per_second": 1.305,
+    "train_loss": 0.3165048904197161,
+    "train_runtime": 4971.9971,
+    "train_samples": 7665,
+    "train_samples_per_second": 3.083,
+    "train_steps_per_second": 0.386
+}

config_argument.yaml

@@ -0,0 +1,47 @@
+cache_dir: ./cache
+ddp_find_unused_parameters: false
+ddp_timeout: 30000
+device_map: auto
+do_eval: true
+do_train: true
+eval_steps: 1000
+evaluation_strategy: steps
+fp16: true
+gradient_accumulation_steps: 1
+gradient_checkpointing: true
+gradient_checkpointing_kwargs:
+  use_reentrant: false
+hub_model_id: hllj/sft-mistral-v3-all
+hub_strategy: every_save
+learning_rate: 5.0e-05
+log_level: info
+logging_first_step: true
+logging_steps: 10
+logging_strategy: steps
+lora_alpha: 128
+lora_dropout: 0.05
+lora_r: 256
+lora_target_modules: all
+lr_scheduler_type: cosine
+max_seq_length: 1024
+model_name_or_path: hllj/mistral-vi-math
+model_type: auto
+num_train_epochs: 2
+output_dir: outputs-sft-mistral-v3-all
+overwrite_output_dir: true
+per_device_eval_batch_size: 8
+per_device_train_batch_size: 8
+preprocessing_num_workers: 4
+push_to_hub: true
+report_to: wandb
+run_name: sft-mistral-v3-all
+save_steps: 1000
+save_strategy: steps
+save_total_limit: 13
+seed: 42
+token: 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
+torch_dtype: float16
+train_file_dir: datasets/finetune
+use_peft: true
+warmup_ratio: 0.05
+weight_decay: 0.05

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.45,
+    "eval_loss": 0.4937315583229065,
+    "eval_runtime": 81.9976,
+    "eval_samples": 852,
+    "eval_samples_per_second": 10.391,
+    "eval_steps_per_second": 1.305
+}

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
+size 493443

tokenizer_config.json

@@ -0,0 +1,43 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [],
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<unk>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": true
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.45,
+    "train_loss": 0.3165048904197161,
+    "train_runtime": 4971.9971,
+    "train_samples": 7665,
+    "train_samples_per_second": 3.083,
+    "train_steps_per_second": 0.386
+}

trainer_state.json

@@ -0,0 +1,550 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.448383733055266,
+  "eval_steps": 1000,
+  "global_step": 860,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.208333333333334e-07,
+      "loss": 0.7516,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.208333333333334e-06,
+      "loss": 0.6842,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 0.6229,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.5266,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.4158,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.604166666666667e-05,
+      "loss": 0.3737,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.125e-05,
+      "loss": 0.3821,
+      "step": 60
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.6458333333333336e-05,
+      "loss": 0.3576,
+      "step": 70
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.351,
+      "step": 80
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.3453,
+      "step": 90
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.999940539127957e-05,
+      "loss": 0.3554,
+      "step": 100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.999271636800334e-05,
+      "loss": 0.3586,
+      "step": 110
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.997859705581399e-05,
+      "loss": 0.3445,
+      "step": 120
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.995705165235725e-05,
+      "loss": 0.3476,
+      "step": 130
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.992808656304221e-05,
+      "loss": 0.33,
+      "step": 140
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9891710399136984e-05,
+      "loss": 0.3354,
+      "step": 150
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.984793397520864e-05,
+      "loss": 0.3356,
+      "step": 160
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9796770305908045e-05,
+      "loss": 0.3216,
+      "step": 170
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.97382346021006e-05,
+      "loss": 0.3353,
+      "step": 180
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.96723442663441e-05,
+      "loss": 0.3216,
+      "step": 190
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.959911888771496e-05,
+      "loss": 0.3251,
+      "step": 200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.951858023598448e-05,
+      "loss": 0.3324,
+      "step": 210
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.943075225514667e-05,
+      "loss": 0.3269,
+      "step": 220
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9335661056299755e-05,
+      "loss": 0.3298,
+      "step": 230
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.923333490988343e-05,
+      "loss": 0.3154,
+      "step": 240
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.912380423727405e-05,
+      "loss": 0.3195,
+      "step": 250
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.900710160174048e-05,
+      "loss": 0.3215,
+      "step": 260
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.888326169876304e-05,
+      "loss": 0.3378,
+      "step": 270
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.875232134571862e-05,
+      "loss": 0.3254,
+      "step": 280
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.8614319470934935e-05,
+      "loss": 0.3513,
+      "step": 290
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.846929710211724e-05,
+      "loss": 0.3355,
+      "step": 300
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.83172973541508e-05,
+      "loss": 0.3238,
+      "step": 310
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.815836541628299e-05,
+      "loss": 0.325,
+      "step": 320
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.7992548538688556e-05,
+      "loss": 0.3256,
+      "step": 330
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.7819896018422237e-05,
+      "loss": 0.3273,
+      "step": 340
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.764045918476288e-05,
+      "loss": 0.3086,
+      "step": 350
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.7454291383953285e-05,
+      "loss": 0.3171,
+      "step": 360
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.726144796334049e-05,
+      "loss": 0.3241,
+      "step": 370
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.706198625492111e-05,
+      "loss": 0.3194,
+      "step": 380
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.685596555829663e-05,
+      "loss": 0.316,
+      "step": 390
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.664344712304375e-05,
+      "loss": 0.315,
+      "step": 400
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.642449413050498e-05,
+      "loss": 0.3209,
+      "step": 410
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.619917167500496e-05,
+      "loss": 0.3146,
+      "step": 420
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.5967546744498046e-05,
+      "loss": 0.3066,
+      "step": 430
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572968820065287e-05,
+      "loss": 0.3024,
+      "step": 440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5485666758379956e-05,
+      "loss": 0.3074,
+      "step": 450
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.5235554964808235e-05,
+      "loss": 0.2973,
+      "step": 460
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.497942717771698e-05,
+      "loss": 0.2799,
+      "step": 470
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.4717359543429316e-05,
+      "loss": 0.291,
+      "step": 480
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.444942997417412e-05,
+      "loss": 0.2803,
+      "step": 490
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.417571812492279e-05,
+      "loss": 0.2679,
+      "step": 500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.389630536970805e-05,
+      "loss": 0.2799,
+      "step": 510
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.361127477743151e-05,
+      "loss": 0.2911,
+      "step": 520
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.332071108716747e-05,
+      "loss": 0.2858,
+      "step": 530
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.302470068297019e-05,
+      "loss": 0.2927,
+      "step": 540
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.2723331568192e-05,
+      "loss": 0.3027,
+      "step": 550
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.2416693339320115e-05,
+      "loss": 0.2884,
+      "step": 560
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.2104877159339726e-05,
+      "loss": 0.305,
+      "step": 570
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.178797573063144e-05,
+      "loss": 0.2814,
+      "step": 580
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.1466083267411005e-05,
+      "loss": 0.2742,
+      "step": 590
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.113929546771963e-05,
+      "loss": 0.2727,
+      "step": 600
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.080770948497311e-05,
+      "loss": 0.3017,
+      "step": 610
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.047142389907827e-05,
+      "loss": 0.2684,
+      "step": 620
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.0130538687125394e-05,
+      "loss": 0.2782,
+      "step": 630
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.978515519366519e-05,
+      "loss": 0.2852,
+      "step": 640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.943537610057921e-05,
+      "loss": 0.2858,
+      "step": 650
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.908130539655278e-05,
+      "loss": 0.2803,
+      "step": 660
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.8723048346159285e-05,
+      "loss": 0.2844,
+      "step": 670
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.836071145856526e-05,
+      "loss": 0.2787,
+      "step": 680
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.7994402455865376e-05,
+      "loss": 0.2719,
+      "step": 690
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.762423024105685e-05,
+      "loss": 0.2798,
+      "step": 700
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.7250304865662857e-05,
+      "loss": 0.2811,
+      "step": 710
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.6872737497014286e-05,
+      "loss": 0.2672,
+      "step": 720
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.649164038520001e-05,
+      "loss": 0.2732,
+      "step": 730
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.6107126829695094e-05,
+      "loss": 0.2759,
+      "step": 740
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.5719311145677055e-05,
+      "loss": 0.2617,
+      "step": 750
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.532830863004018e-05,
+      "loss": 0.2809,
+      "step": 760
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.4934235527118e-05,
+      "loss": 0.2662,
+      "step": 770
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.4537208994124014e-05,
+      "loss": 0.2496,
+      "step": 780
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.41373470663211e-05,
+      "loss": 0.2761,
+      "step": 790
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.37347686219298e-05,
+      "loss": 0.2621,
+      "step": 800
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.3329593346786125e-05,
+      "loss": 0.264,
+      "step": 810
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.292194169875908e-05,
+      "loss": 0.2552,
+      "step": 820
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.251193487193883e-05,
+      "loss": 0.2675,
+      "step": 830
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.209969476060587e-05,
+      "loss": 0.2614,
+      "step": 840
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.168534392299214e-05,
+      "loss": 0.2721,
+      "step": 850
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.126900554484459e-05,
+      "loss": 0.2751,
+      "step": 860
+    },
+    {
+      "epoch": 1.45,
+      "step": 860,
+      "total_flos": 3.091617472671908e+17,
+      "train_loss": 0.3165048904197161,
+      "train_runtime": 4971.9971,
+      "train_samples_per_second": 3.083,
+      "train_steps_per_second": 0.386
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 1918,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "total_flos": 3.091617472671908e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c3afe54135ad942380dee2eef4388fa893295c951f98e6eb91b191f536cecc4
+size 4664