Upload folder using huggingface_hub
Browse files- .gitattributes +4 -0
- checkpoint-300/adapter_config.json +42 -0
- checkpoint-300/special_tokens_map.json +24 -0
- checkpoint-300/tokenizer.json +3 -0
- checkpoint-300/tokenizer_config.json +0 -0
- checkpoint-300/trainer_state.json +400 -0
- checkpoint-400/adapter_config.json +42 -0
- checkpoint-400/special_tokens_map.json +24 -0
- checkpoint-400/tokenizer.json +3 -0
- checkpoint-400/tokenizer_config.json +0 -0
- checkpoint-400/trainer_state.json +522 -0
- checkpoint-500/adapter_config.json +42 -0
- checkpoint-500/special_tokens_map.json +24 -0
- checkpoint-500/tokenizer.json +3 -0
- checkpoint-500/tokenizer_config.json +0 -0
- checkpoint-500/trainer_state.json +644 -0
- special_tokens_map.json +24 -0
- tokenizer.json +3 -0
- tokenizer_config.json +0 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
checkpoint-300/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
checkpoint-400/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
checkpoint-500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
checkpoint-300/adapter_config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "/home/support/llm/Ministral-8B-Instruct-2410/",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"corda_config": null,
|
| 7 |
+
"eva_config": null,
|
| 8 |
+
"exclude_modules": null,
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layer_replication": null,
|
| 13 |
+
"layers_pattern": null,
|
| 14 |
+
"layers_to_transform": null,
|
| 15 |
+
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 16,
|
| 17 |
+
"lora_bias": false,
|
| 18 |
+
"lora_dropout": 0.05,
|
| 19 |
+
"megatron_config": null,
|
| 20 |
+
"megatron_core": "megatron.core",
|
| 21 |
+
"modules_to_save": null,
|
| 22 |
+
"peft_type": "LORA",
|
| 23 |
+
"qalora_group_size": 16,
|
| 24 |
+
"r": 64,
|
| 25 |
+
"rank_pattern": {},
|
| 26 |
+
"revision": null,
|
| 27 |
+
"target_modules": [
|
| 28 |
+
"up_proj",
|
| 29 |
+
"down_proj",
|
| 30 |
+
"k_proj",
|
| 31 |
+
"gate_proj",
|
| 32 |
+
"v_proj",
|
| 33 |
+
"q_proj",
|
| 34 |
+
"o_proj"
|
| 35 |
+
],
|
| 36 |
+
"target_parameters": null,
|
| 37 |
+
"task_type": "CAUSAL_LM",
|
| 38 |
+
"trainable_token_indices": null,
|
| 39 |
+
"use_dora": false,
|
| 40 |
+
"use_qalora": false,
|
| 41 |
+
"use_rslora": false
|
| 42 |
+
}
|
checkpoint-300/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-300/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7edbeaf20dd7f571b5dd1c54d9ace4f9b6299127cc7ba2afb14a6d51a4a79a4
|
| 3 |
+
size 17078136
|
checkpoint-300/tokenizer_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-300/trainer_state.json
ADDED
|
@@ -0,0 +1,400 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 300,
|
| 3 |
+
"best_metric": 0.5047101974487305,
|
| 4 |
+
"best_model_checkpoint": "./python-code-cot-ministral-8B/checkpoint-300",
|
| 5 |
+
"epoch": 1.2940540540540542,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 300,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 0.8707441985607147,
|
| 14 |
+
"epoch": 0.043243243243243246,
|
| 15 |
+
"grad_norm": 0.2866508960723877,
|
| 16 |
+
"learning_rate": 6e-06,
|
| 17 |
+
"loss": 1.065,
|
| 18 |
+
"mean_token_accuracy": 0.7471851870417595,
|
| 19 |
+
"num_tokens": 40787.0,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 0.818158182501793,
|
| 24 |
+
"epoch": 0.08648648648648649,
|
| 25 |
+
"grad_norm": 0.2426566779613495,
|
| 26 |
+
"learning_rate": 9.998321768134101e-06,
|
| 27 |
+
"loss": 0.9899,
|
| 28 |
+
"mean_token_accuracy": 0.7604838714003563,
|
| 29 |
+
"num_tokens": 81747.0,
|
| 30 |
+
"step": 20
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 0.8590745508670807,
|
| 34 |
+
"epoch": 0.12972972972972974,
|
| 35 |
+
"grad_norm": 0.2603059709072113,
|
| 36 |
+
"learning_rate": 9.979454595377594e-06,
|
| 37 |
+
"loss": 0.9586,
|
| 38 |
+
"mean_token_accuracy": 0.7590420320630074,
|
| 39 |
+
"num_tokens": 122707.0,
|
| 40 |
+
"step": 30
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 0.7979068487882615,
|
| 44 |
+
"epoch": 0.17297297297297298,
|
| 45 |
+
"grad_norm": 0.2751409411430359,
|
| 46 |
+
"learning_rate": 9.93970185956522e-06,
|
| 47 |
+
"loss": 0.809,
|
| 48 |
+
"mean_token_accuracy": 0.7851744830608368,
|
| 49 |
+
"num_tokens": 163305.0,
|
| 50 |
+
"step": 40
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"entropy": 0.7995434030890465,
|
| 54 |
+
"epoch": 0.21621621621621623,
|
| 55 |
+
"grad_norm": 0.27979347109794617,
|
| 56 |
+
"learning_rate": 9.879230297486034e-06,
|
| 57 |
+
"loss": 0.7849,
|
| 58 |
+
"mean_token_accuracy": 0.791813300549984,
|
| 59 |
+
"num_tokens": 204265.0,
|
| 60 |
+
"step": 50
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.21621621621621623,
|
| 64 |
+
"eval_entropy": 0.7247167360491868,
|
| 65 |
+
"eval_loss": 0.7107394337654114,
|
| 66 |
+
"eval_mean_token_accuracy": 0.813098669052124,
|
| 67 |
+
"eval_num_tokens": 204265.0,
|
| 68 |
+
"eval_runtime": 40.5827,
|
| 69 |
+
"eval_samples_per_second": 4.041,
|
| 70 |
+
"eval_steps_per_second": 4.041,
|
| 71 |
+
"step": 50
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"entropy": 0.6114772140979767,
|
| 75 |
+
"epoch": 0.2594594594594595,
|
| 76 |
+
"grad_norm": 0.1919497400522232,
|
| 77 |
+
"learning_rate": 9.798293547886748e-06,
|
| 78 |
+
"loss": 0.6123,
|
| 79 |
+
"mean_token_accuracy": 0.8428526982665062,
|
| 80 |
+
"num_tokens": 245044.0,
|
| 81 |
+
"step": 60
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"entropy": 0.5336497366428375,
|
| 85 |
+
"epoch": 0.3027027027027027,
|
| 86 |
+
"grad_norm": 0.20059765875339508,
|
| 87 |
+
"learning_rate": 9.697231087622691e-06,
|
| 88 |
+
"loss": 0.5426,
|
| 89 |
+
"mean_token_accuracy": 0.8581829935312271,
|
| 90 |
+
"num_tokens": 285929.0,
|
| 91 |
+
"step": 70
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"entropy": 0.49029816724359987,
|
| 95 |
+
"epoch": 0.34594594594594597,
|
| 96 |
+
"grad_norm": 0.19071193039417267,
|
| 97 |
+
"learning_rate": 9.5764668077739e-06,
|
| 98 |
+
"loss": 0.5078,
|
| 99 |
+
"mean_token_accuracy": 0.8675507545471192,
|
| 100 |
+
"num_tokens": 326731.0,
|
| 101 |
+
"step": 80
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"entropy": 0.568328783661127,
|
| 105 |
+
"epoch": 0.3891891891891892,
|
| 106 |
+
"grad_norm": 0.1908392310142517,
|
| 107 |
+
"learning_rate": 9.436507235698613e-06,
|
| 108 |
+
"loss": 0.5825,
|
| 109 |
+
"mean_token_accuracy": 0.8428896099328995,
|
| 110 |
+
"num_tokens": 367252.0,
|
| 111 |
+
"step": 90
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"entropy": 0.5580322951078415,
|
| 115 |
+
"epoch": 0.43243243243243246,
|
| 116 |
+
"grad_norm": 0.1762734353542328,
|
| 117 |
+
"learning_rate": 9.277939410481507e-06,
|
| 118 |
+
"loss": 0.5536,
|
| 119 |
+
"mean_token_accuracy": 0.8481671541929245,
|
| 120 |
+
"num_tokens": 408212.0,
|
| 121 |
+
"step": 100
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.43243243243243246,
|
| 125 |
+
"eval_entropy": 0.5617244329939528,
|
| 126 |
+
"eval_loss": 0.5632628202438354,
|
| 127 |
+
"eval_mean_token_accuracy": 0.8465738710833759,
|
| 128 |
+
"eval_num_tokens": 408212.0,
|
| 129 |
+
"eval_runtime": 40.5739,
|
| 130 |
+
"eval_samples_per_second": 4.042,
|
| 131 |
+
"eval_steps_per_second": 4.042,
|
| 132 |
+
"step": 100
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 0.5392510816454887,
|
| 136 |
+
"epoch": 0.4756756756756757,
|
| 137 |
+
"grad_norm": 0.13735364377498627,
|
| 138 |
+
"learning_rate": 9.101428420687759e-06,
|
| 139 |
+
"loss": 0.5391,
|
| 140 |
+
"mean_token_accuracy": 0.8521016612648964,
|
| 141 |
+
"num_tokens": 449172.0,
|
| 142 |
+
"step": 110
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 0.6197883028537035,
|
| 146 |
+
"epoch": 0.518918918918919,
|
| 147 |
+
"grad_norm": 0.1790686547756195,
|
| 148 |
+
"learning_rate": 8.907714614750473e-06,
|
| 149 |
+
"loss": 0.6066,
|
| 150 |
+
"mean_token_accuracy": 0.837218964099884,
|
| 151 |
+
"num_tokens": 490132.0,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"entropy": 0.5101678855717182,
|
| 156 |
+
"epoch": 0.5621621621621622,
|
| 157 |
+
"grad_norm": 0.16557781398296356,
|
| 158 |
+
"learning_rate": 8.697610495692055e-06,
|
| 159 |
+
"loss": 0.4937,
|
| 160 |
+
"mean_token_accuracy": 0.861264032125473,
|
| 161 |
+
"num_tokens": 531046.0,
|
| 162 |
+
"step": 130
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"entropy": 0.5256278648972511,
|
| 166 |
+
"epoch": 0.6054054054054054,
|
| 167 |
+
"grad_norm": 0.15723901987075806,
|
| 168 |
+
"learning_rate": 8.471997313204183e-06,
|
| 169 |
+
"loss": 0.5128,
|
| 170 |
+
"mean_token_accuracy": 0.8562316685914994,
|
| 171 |
+
"num_tokens": 572006.0,
|
| 172 |
+
"step": 140
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"entropy": 0.5276982985436917,
|
| 176 |
+
"epoch": 0.6486486486486487,
|
| 177 |
+
"grad_norm": 0.14954417943954468,
|
| 178 |
+
"learning_rate": 8.231821367380335e-06,
|
| 179 |
+
"loss": 0.5148,
|
| 180 |
+
"mean_token_accuracy": 0.8543255120515824,
|
| 181 |
+
"num_tokens": 612966.0,
|
| 182 |
+
"step": 150
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.6486486486486487,
|
| 186 |
+
"eval_entropy": 0.5507546397970944,
|
| 187 |
+
"eval_loss": 0.5337628722190857,
|
| 188 |
+
"eval_mean_token_accuracy": 0.8515133745059734,
|
| 189 |
+
"eval_num_tokens": 612966.0,
|
| 190 |
+
"eval_runtime": 40.52,
|
| 191 |
+
"eval_samples_per_second": 4.047,
|
| 192 |
+
"eval_steps_per_second": 4.047,
|
| 193 |
+
"step": 150
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"entropy": 0.5230918094515801,
|
| 197 |
+
"epoch": 0.6918918918918919,
|
| 198 |
+
"grad_norm": 0.14227904379367828,
|
| 199 |
+
"learning_rate": 7.978090039604342e-06,
|
| 200 |
+
"loss": 0.5118,
|
| 201 |
+
"mean_token_accuracy": 0.8542912498116493,
|
| 202 |
+
"num_tokens": 653680.0,
|
| 203 |
+
"step": 160
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"entropy": 0.524362176656723,
|
| 207 |
+
"epoch": 0.7351351351351352,
|
| 208 |
+
"grad_norm": 0.16846755146980286,
|
| 209 |
+
"learning_rate": 7.711867567242769e-06,
|
| 210 |
+
"loss": 0.5104,
|
| 211 |
+
"mean_token_accuracy": 0.8555971384048462,
|
| 212 |
+
"num_tokens": 694478.0,
|
| 213 |
+
"step": 170
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"entropy": 0.49619704745709897,
|
| 217 |
+
"epoch": 0.7783783783783784,
|
| 218 |
+
"grad_norm": 0.20149804651737213,
|
| 219 |
+
"learning_rate": 7.434270579863549e-06,
|
| 220 |
+
"loss": 0.4826,
|
| 221 |
+
"mean_token_accuracy": 0.8604745492339134,
|
| 222 |
+
"num_tokens": 735351.0,
|
| 223 |
+
"step": 180
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"entropy": 0.5439279317855835,
|
| 227 |
+
"epoch": 0.8216216216216217,
|
| 228 |
+
"grad_norm": 0.22333768010139465,
|
| 229 |
+
"learning_rate": 7.14646341570353e-06,
|
| 230 |
+
"loss": 0.5411,
|
| 231 |
+
"mean_token_accuracy": 0.8482905030250549,
|
| 232 |
+
"num_tokens": 775651.0,
|
| 233 |
+
"step": 190
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"entropy": 0.4976503163576126,
|
| 237 |
+
"epoch": 0.8648648648648649,
|
| 238 |
+
"grad_norm": 0.16309203207492828,
|
| 239 |
+
"learning_rate": 6.849653238029261e-06,
|
| 240 |
+
"loss": 0.4994,
|
| 241 |
+
"mean_token_accuracy": 0.85667155534029,
|
| 242 |
+
"num_tokens": 816611.0,
|
| 243 |
+
"step": 200
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.8648648648648649,
|
| 247 |
+
"eval_entropy": 0.5280105926823325,
|
| 248 |
+
"eval_loss": 0.5184423327445984,
|
| 249 |
+
"eval_mean_token_accuracy": 0.8534702516910506,
|
| 250 |
+
"eval_num_tokens": 816611.0,
|
| 251 |
+
"eval_runtime": 40.6179,
|
| 252 |
+
"eval_samples_per_second": 4.038,
|
| 253 |
+
"eval_steps_per_second": 4.038,
|
| 254 |
+
"step": 200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"entropy": 0.5327252022922039,
|
| 258 |
+
"epoch": 0.9081081081081082,
|
| 259 |
+
"grad_norm": 0.16586290299892426,
|
| 260 |
+
"learning_rate": 6.545084971874738e-06,
|
| 261 |
+
"loss": 0.5268,
|
| 262 |
+
"mean_token_accuracy": 0.8529814258217812,
|
| 263 |
+
"num_tokens": 857571.0,
|
| 264 |
+
"step": 210
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"entropy": 0.47181336022913456,
|
| 268 |
+
"epoch": 0.9513513513513514,
|
| 269 |
+
"grad_norm": 0.17959456145763397,
|
| 270 |
+
"learning_rate": 6.234036082393171e-06,
|
| 271 |
+
"loss": 0.4687,
|
| 272 |
+
"mean_token_accuracy": 0.8678423285484314,
|
| 273 |
+
"num_tokens": 898415.0,
|
| 274 |
+
"step": 220
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"entropy": 0.5040528390556573,
|
| 278 |
+
"epoch": 0.9945945945945946,
|
| 279 |
+
"grad_norm": 0.17449511587619781,
|
| 280 |
+
"learning_rate": 5.9178112167241805e-06,
|
| 281 |
+
"loss": 0.4999,
|
| 282 |
+
"mean_token_accuracy": 0.8588241055607796,
|
| 283 |
+
"num_tokens": 939275.0,
|
| 284 |
+
"step": 230
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"entropy": 0.4610033663543495,
|
| 288 |
+
"epoch": 1.0345945945945947,
|
| 289 |
+
"grad_norm": 0.15696631371974945,
|
| 290 |
+
"learning_rate": 5.597736731850295e-06,
|
| 291 |
+
"loss": 0.4834,
|
| 292 |
+
"mean_token_accuracy": 0.8710998361175125,
|
| 293 |
+
"num_tokens": 977163.0,
|
| 294 |
+
"step": 240
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"entropy": 0.5211307317018509,
|
| 298 |
+
"epoch": 1.077837837837838,
|
| 299 |
+
"grad_norm": 0.1736782193183899,
|
| 300 |
+
"learning_rate": 5.275155131394825e-06,
|
| 301 |
+
"loss": 0.5047,
|
| 302 |
+
"mean_token_accuracy": 0.854168564081192,
|
| 303 |
+
"num_tokens": 1017950.0,
|
| 304 |
+
"step": 250
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.077837837837838,
|
| 308 |
+
"eval_entropy": 0.5199673544706368,
|
| 309 |
+
"eval_loss": 0.5092993974685669,
|
| 310 |
+
"eval_mean_token_accuracy": 0.855072995511497,
|
| 311 |
+
"eval_num_tokens": 1017950.0,
|
| 312 |
+
"eval_runtime": 40.6579,
|
| 313 |
+
"eval_samples_per_second": 4.034,
|
| 314 |
+
"eval_steps_per_second": 4.034,
|
| 315 |
+
"step": 250
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"entropy": 0.4647901371121407,
|
| 319 |
+
"epoch": 1.1210810810810812,
|
| 320 |
+
"grad_norm": 0.18243983387947083,
|
| 321 |
+
"learning_rate": 4.951419434695115e-06,
|
| 322 |
+
"loss": 0.4514,
|
| 323 |
+
"mean_token_accuracy": 0.8703323498368263,
|
| 324 |
+
"num_tokens": 1058910.0,
|
| 325 |
+
"step": 260
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"entropy": 0.4522815000265837,
|
| 329 |
+
"epoch": 1.1643243243243244,
|
| 330 |
+
"grad_norm": 0.20347489416599274,
|
| 331 |
+
"learning_rate": 4.627887501769231e-06,
|
| 332 |
+
"loss": 0.4454,
|
| 333 |
+
"mean_token_accuracy": 0.8707166478037834,
|
| 334 |
+
"num_tokens": 1099604.0,
|
| 335 |
+
"step": 270
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"entropy": 0.4927990548312664,
|
| 339 |
+
"epoch": 1.2075675675675677,
|
| 340 |
+
"grad_norm": 0.18338286876678467,
|
| 341 |
+
"learning_rate": 4.3059163379791676e-06,
|
| 342 |
+
"loss": 0.491,
|
| 343 |
+
"mean_token_accuracy": 0.8596193596720696,
|
| 344 |
+
"num_tokens": 1140464.0,
|
| 345 |
+
"step": 280
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"entropy": 0.4583901420235634,
|
| 349 |
+
"epoch": 1.250810810810811,
|
| 350 |
+
"grad_norm": 0.18116827309131622,
|
| 351 |
+
"learning_rate": 3.98685640227875e-06,
|
| 352 |
+
"loss": 0.4446,
|
| 353 |
+
"mean_token_accuracy": 0.8701988190412522,
|
| 354 |
+
"num_tokens": 1181280.0,
|
| 355 |
+
"step": 290
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"entropy": 0.5115556083619595,
|
| 359 |
+
"epoch": 1.2940540540540542,
|
| 360 |
+
"grad_norm": 0.21102362871170044,
|
| 361 |
+
"learning_rate": 3.6720459429194743e-06,
|
| 362 |
+
"loss": 0.5013,
|
| 363 |
+
"mean_token_accuracy": 0.8555837154388428,
|
| 364 |
+
"num_tokens": 1221884.0,
|
| 365 |
+
"step": 300
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.2940540540540542,
|
| 369 |
+
"eval_entropy": 0.5066255062091641,
|
| 370 |
+
"eval_loss": 0.5047101974487305,
|
| 371 |
+
"eval_mean_token_accuracy": 0.856273435964817,
|
| 372 |
+
"eval_num_tokens": 1221884.0,
|
| 373 |
+
"eval_runtime": 40.5902,
|
| 374 |
+
"eval_samples_per_second": 4.04,
|
| 375 |
+
"eval_steps_per_second": 4.04,
|
| 376 |
+
"step": 300
|
| 377 |
+
}
|
| 378 |
+
],
|
| 379 |
+
"logging_steps": 10,
|
| 380 |
+
"max_steps": 500,
|
| 381 |
+
"num_input_tokens_seen": 0,
|
| 382 |
+
"num_train_epochs": 3,
|
| 383 |
+
"save_steps": 100,
|
| 384 |
+
"stateful_callbacks": {
|
| 385 |
+
"TrainerControl": {
|
| 386 |
+
"args": {
|
| 387 |
+
"should_epoch_stop": false,
|
| 388 |
+
"should_evaluate": false,
|
| 389 |
+
"should_log": false,
|
| 390 |
+
"should_save": true,
|
| 391 |
+
"should_training_stop": false
|
| 392 |
+
},
|
| 393 |
+
"attributes": {}
|
| 394 |
+
}
|
| 395 |
+
},
|
| 396 |
+
"total_flos": 5.613964548076339e+16,
|
| 397 |
+
"train_batch_size": 1,
|
| 398 |
+
"trial_name": null,
|
| 399 |
+
"trial_params": null
|
| 400 |
+
}
|
checkpoint-400/adapter_config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "/home/support/llm/Ministral-8B-Instruct-2410/",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"corda_config": null,
|
| 7 |
+
"eva_config": null,
|
| 8 |
+
"exclude_modules": null,
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layer_replication": null,
|
| 13 |
+
"layers_pattern": null,
|
| 14 |
+
"layers_to_transform": null,
|
| 15 |
+
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 16,
|
| 17 |
+
"lora_bias": false,
|
| 18 |
+
"lora_dropout": 0.05,
|
| 19 |
+
"megatron_config": null,
|
| 20 |
+
"megatron_core": "megatron.core",
|
| 21 |
+
"modules_to_save": null,
|
| 22 |
+
"peft_type": "LORA",
|
| 23 |
+
"qalora_group_size": 16,
|
| 24 |
+
"r": 64,
|
| 25 |
+
"rank_pattern": {},
|
| 26 |
+
"revision": null,
|
| 27 |
+
"target_modules": [
|
| 28 |
+
"up_proj",
|
| 29 |
+
"down_proj",
|
| 30 |
+
"k_proj",
|
| 31 |
+
"gate_proj",
|
| 32 |
+
"v_proj",
|
| 33 |
+
"q_proj",
|
| 34 |
+
"o_proj"
|
| 35 |
+
],
|
| 36 |
+
"target_parameters": null,
|
| 37 |
+
"task_type": "CAUSAL_LM",
|
| 38 |
+
"trainable_token_indices": null,
|
| 39 |
+
"use_dora": false,
|
| 40 |
+
"use_qalora": false,
|
| 41 |
+
"use_rslora": false
|
| 42 |
+
}
|
checkpoint-400/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-400/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7edbeaf20dd7f571b5dd1c54d9ace4f9b6299127cc7ba2afb14a6d51a4a79a4
|
| 3 |
+
size 17078136
|
checkpoint-400/tokenizer_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-400/trainer_state.json
ADDED
|
@@ -0,0 +1,522 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 400,
|
| 3 |
+
"best_metric": 0.5008729100227356,
|
| 4 |
+
"best_model_checkpoint": "./python-code-cot-ministral-8B/checkpoint-400",
|
| 5 |
+
"epoch": 1.7264864864864866,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 400,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 0.8707441985607147,
|
| 14 |
+
"epoch": 0.043243243243243246,
|
| 15 |
+
"grad_norm": 0.2866508960723877,
|
| 16 |
+
"learning_rate": 6e-06,
|
| 17 |
+
"loss": 1.065,
|
| 18 |
+
"mean_token_accuracy": 0.7471851870417595,
|
| 19 |
+
"num_tokens": 40787.0,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 0.818158182501793,
|
| 24 |
+
"epoch": 0.08648648648648649,
|
| 25 |
+
"grad_norm": 0.2426566779613495,
|
| 26 |
+
"learning_rate": 9.998321768134101e-06,
|
| 27 |
+
"loss": 0.9899,
|
| 28 |
+
"mean_token_accuracy": 0.7604838714003563,
|
| 29 |
+
"num_tokens": 81747.0,
|
| 30 |
+
"step": 20
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 0.8590745508670807,
|
| 34 |
+
"epoch": 0.12972972972972974,
|
| 35 |
+
"grad_norm": 0.2603059709072113,
|
| 36 |
+
"learning_rate": 9.979454595377594e-06,
|
| 37 |
+
"loss": 0.9586,
|
| 38 |
+
"mean_token_accuracy": 0.7590420320630074,
|
| 39 |
+
"num_tokens": 122707.0,
|
| 40 |
+
"step": 30
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 0.7979068487882615,
|
| 44 |
+
"epoch": 0.17297297297297298,
|
| 45 |
+
"grad_norm": 0.2751409411430359,
|
| 46 |
+
"learning_rate": 9.93970185956522e-06,
|
| 47 |
+
"loss": 0.809,
|
| 48 |
+
"mean_token_accuracy": 0.7851744830608368,
|
| 49 |
+
"num_tokens": 163305.0,
|
| 50 |
+
"step": 40
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"entropy": 0.7995434030890465,
|
| 54 |
+
"epoch": 0.21621621621621623,
|
| 55 |
+
"grad_norm": 0.27979347109794617,
|
| 56 |
+
"learning_rate": 9.879230297486034e-06,
|
| 57 |
+
"loss": 0.7849,
|
| 58 |
+
"mean_token_accuracy": 0.791813300549984,
|
| 59 |
+
"num_tokens": 204265.0,
|
| 60 |
+
"step": 50
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.21621621621621623,
|
| 64 |
+
"eval_entropy": 0.7247167360491868,
|
| 65 |
+
"eval_loss": 0.7107394337654114,
|
| 66 |
+
"eval_mean_token_accuracy": 0.813098669052124,
|
| 67 |
+
"eval_num_tokens": 204265.0,
|
| 68 |
+
"eval_runtime": 40.5827,
|
| 69 |
+
"eval_samples_per_second": 4.041,
|
| 70 |
+
"eval_steps_per_second": 4.041,
|
| 71 |
+
"step": 50
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"entropy": 0.6114772140979767,
|
| 75 |
+
"epoch": 0.2594594594594595,
|
| 76 |
+
"grad_norm": 0.1919497400522232,
|
| 77 |
+
"learning_rate": 9.798293547886748e-06,
|
| 78 |
+
"loss": 0.6123,
|
| 79 |
+
"mean_token_accuracy": 0.8428526982665062,
|
| 80 |
+
"num_tokens": 245044.0,
|
| 81 |
+
"step": 60
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"entropy": 0.5336497366428375,
|
| 85 |
+
"epoch": 0.3027027027027027,
|
| 86 |
+
"grad_norm": 0.20059765875339508,
|
| 87 |
+
"learning_rate": 9.697231087622691e-06,
|
| 88 |
+
"loss": 0.5426,
|
| 89 |
+
"mean_token_accuracy": 0.8581829935312271,
|
| 90 |
+
"num_tokens": 285929.0,
|
| 91 |
+
"step": 70
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"entropy": 0.49029816724359987,
|
| 95 |
+
"epoch": 0.34594594594594597,
|
| 96 |
+
"grad_norm": 0.19071193039417267,
|
| 97 |
+
"learning_rate": 9.5764668077739e-06,
|
| 98 |
+
"loss": 0.5078,
|
| 99 |
+
"mean_token_accuracy": 0.8675507545471192,
|
| 100 |
+
"num_tokens": 326731.0,
|
| 101 |
+
"step": 80
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"entropy": 0.568328783661127,
|
| 105 |
+
"epoch": 0.3891891891891892,
|
| 106 |
+
"grad_norm": 0.1908392310142517,
|
| 107 |
+
"learning_rate": 9.436507235698613e-06,
|
| 108 |
+
"loss": 0.5825,
|
| 109 |
+
"mean_token_accuracy": 0.8428896099328995,
|
| 110 |
+
"num_tokens": 367252.0,
|
| 111 |
+
"step": 90
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"entropy": 0.5580322951078415,
|
| 115 |
+
"epoch": 0.43243243243243246,
|
| 116 |
+
"grad_norm": 0.1762734353542328,
|
| 117 |
+
"learning_rate": 9.277939410481507e-06,
|
| 118 |
+
"loss": 0.5536,
|
| 119 |
+
"mean_token_accuracy": 0.8481671541929245,
|
| 120 |
+
"num_tokens": 408212.0,
|
| 121 |
+
"step": 100
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.43243243243243246,
|
| 125 |
+
"eval_entropy": 0.5617244329939528,
|
| 126 |
+
"eval_loss": 0.5632628202438354,
|
| 127 |
+
"eval_mean_token_accuracy": 0.8465738710833759,
|
| 128 |
+
"eval_num_tokens": 408212.0,
|
| 129 |
+
"eval_runtime": 40.5739,
|
| 130 |
+
"eval_samples_per_second": 4.042,
|
| 131 |
+
"eval_steps_per_second": 4.042,
|
| 132 |
+
"step": 100
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 0.5392510816454887,
|
| 136 |
+
"epoch": 0.4756756756756757,
|
| 137 |
+
"grad_norm": 0.13735364377498627,
|
| 138 |
+
"learning_rate": 9.101428420687759e-06,
|
| 139 |
+
"loss": 0.5391,
|
| 140 |
+
"mean_token_accuracy": 0.8521016612648964,
|
| 141 |
+
"num_tokens": 449172.0,
|
| 142 |
+
"step": 110
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 0.6197883028537035,
|
| 146 |
+
"epoch": 0.518918918918919,
|
| 147 |
+
"grad_norm": 0.1790686547756195,
|
| 148 |
+
"learning_rate": 8.907714614750473e-06,
|
| 149 |
+
"loss": 0.6066,
|
| 150 |
+
"mean_token_accuracy": 0.837218964099884,
|
| 151 |
+
"num_tokens": 490132.0,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"entropy": 0.5101678855717182,
|
| 156 |
+
"epoch": 0.5621621621621622,
|
| 157 |
+
"grad_norm": 0.16557781398296356,
|
| 158 |
+
"learning_rate": 8.697610495692055e-06,
|
| 159 |
+
"loss": 0.4937,
|
| 160 |
+
"mean_token_accuracy": 0.861264032125473,
|
| 161 |
+
"num_tokens": 531046.0,
|
| 162 |
+
"step": 130
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"entropy": 0.5256278648972511,
|
| 166 |
+
"epoch": 0.6054054054054054,
|
| 167 |
+
"grad_norm": 0.15723901987075806,
|
| 168 |
+
"learning_rate": 8.471997313204183e-06,
|
| 169 |
+
"loss": 0.5128,
|
| 170 |
+
"mean_token_accuracy": 0.8562316685914994,
|
| 171 |
+
"num_tokens": 572006.0,
|
| 172 |
+
"step": 140
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"entropy": 0.5276982985436917,
|
| 176 |
+
"epoch": 0.6486486486486487,
|
| 177 |
+
"grad_norm": 0.14954417943954468,
|
| 178 |
+
"learning_rate": 8.231821367380335e-06,
|
| 179 |
+
"loss": 0.5148,
|
| 180 |
+
"mean_token_accuracy": 0.8543255120515824,
|
| 181 |
+
"num_tokens": 612966.0,
|
| 182 |
+
"step": 150
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.6486486486486487,
|
| 186 |
+
"eval_entropy": 0.5507546397970944,
|
| 187 |
+
"eval_loss": 0.5337628722190857,
|
| 188 |
+
"eval_mean_token_accuracy": 0.8515133745059734,
|
| 189 |
+
"eval_num_tokens": 612966.0,
|
| 190 |
+
"eval_runtime": 40.52,
|
| 191 |
+
"eval_samples_per_second": 4.047,
|
| 192 |
+
"eval_steps_per_second": 4.047,
|
| 193 |
+
"step": 150
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"entropy": 0.5230918094515801,
|
| 197 |
+
"epoch": 0.6918918918918919,
|
| 198 |
+
"grad_norm": 0.14227904379367828,
|
| 199 |
+
"learning_rate": 7.978090039604342e-06,
|
| 200 |
+
"loss": 0.5118,
|
| 201 |
+
"mean_token_accuracy": 0.8542912498116493,
|
| 202 |
+
"num_tokens": 653680.0,
|
| 203 |
+
"step": 160
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"entropy": 0.524362176656723,
|
| 207 |
+
"epoch": 0.7351351351351352,
|
| 208 |
+
"grad_norm": 0.16846755146980286,
|
| 209 |
+
"learning_rate": 7.711867567242769e-06,
|
| 210 |
+
"loss": 0.5104,
|
| 211 |
+
"mean_token_accuracy": 0.8555971384048462,
|
| 212 |
+
"num_tokens": 694478.0,
|
| 213 |
+
"step": 170
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"entropy": 0.49619704745709897,
|
| 217 |
+
"epoch": 0.7783783783783784,
|
| 218 |
+
"grad_norm": 0.20149804651737213,
|
| 219 |
+
"learning_rate": 7.434270579863549e-06,
|
| 220 |
+
"loss": 0.4826,
|
| 221 |
+
"mean_token_accuracy": 0.8604745492339134,
|
| 222 |
+
"num_tokens": 735351.0,
|
| 223 |
+
"step": 180
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"entropy": 0.5439279317855835,
|
| 227 |
+
"epoch": 0.8216216216216217,
|
| 228 |
+
"grad_norm": 0.22333768010139465,
|
| 229 |
+
"learning_rate": 7.14646341570353e-06,
|
| 230 |
+
"loss": 0.5411,
|
| 231 |
+
"mean_token_accuracy": 0.8482905030250549,
|
| 232 |
+
"num_tokens": 775651.0,
|
| 233 |
+
"step": 190
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"entropy": 0.4976503163576126,
|
| 237 |
+
"epoch": 0.8648648648648649,
|
| 238 |
+
"grad_norm": 0.16309203207492828,
|
| 239 |
+
"learning_rate": 6.849653238029261e-06,
|
| 240 |
+
"loss": 0.4994,
|
| 241 |
+
"mean_token_accuracy": 0.85667155534029,
|
| 242 |
+
"num_tokens": 816611.0,
|
| 243 |
+
"step": 200
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.8648648648648649,
|
| 247 |
+
"eval_entropy": 0.5280105926823325,
|
| 248 |
+
"eval_loss": 0.5184423327445984,
|
| 249 |
+
"eval_mean_token_accuracy": 0.8534702516910506,
|
| 250 |
+
"eval_num_tokens": 816611.0,
|
| 251 |
+
"eval_runtime": 40.6179,
|
| 252 |
+
"eval_samples_per_second": 4.038,
|
| 253 |
+
"eval_steps_per_second": 4.038,
|
| 254 |
+
"step": 200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"entropy": 0.5327252022922039,
|
| 258 |
+
"epoch": 0.9081081081081082,
|
| 259 |
+
"grad_norm": 0.16586290299892426,
|
| 260 |
+
"learning_rate": 6.545084971874738e-06,
|
| 261 |
+
"loss": 0.5268,
|
| 262 |
+
"mean_token_accuracy": 0.8529814258217812,
|
| 263 |
+
"num_tokens": 857571.0,
|
| 264 |
+
"step": 210
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"entropy": 0.47181336022913456,
|
| 268 |
+
"epoch": 0.9513513513513514,
|
| 269 |
+
"grad_norm": 0.17959456145763397,
|
| 270 |
+
"learning_rate": 6.234036082393171e-06,
|
| 271 |
+
"loss": 0.4687,
|
| 272 |
+
"mean_token_accuracy": 0.8678423285484314,
|
| 273 |
+
"num_tokens": 898415.0,
|
| 274 |
+
"step": 220
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"entropy": 0.5040528390556573,
|
| 278 |
+
"epoch": 0.9945945945945946,
|
| 279 |
+
"grad_norm": 0.17449511587619781,
|
| 280 |
+
"learning_rate": 5.9178112167241805e-06,
|
| 281 |
+
"loss": 0.4999,
|
| 282 |
+
"mean_token_accuracy": 0.8588241055607796,
|
| 283 |
+
"num_tokens": 939275.0,
|
| 284 |
+
"step": 230
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"entropy": 0.4610033663543495,
|
| 288 |
+
"epoch": 1.0345945945945947,
|
| 289 |
+
"grad_norm": 0.15696631371974945,
|
| 290 |
+
"learning_rate": 5.597736731850295e-06,
|
| 291 |
+
"loss": 0.4834,
|
| 292 |
+
"mean_token_accuracy": 0.8710998361175125,
|
| 293 |
+
"num_tokens": 977163.0,
|
| 294 |
+
"step": 240
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"entropy": 0.5211307317018509,
|
| 298 |
+
"epoch": 1.077837837837838,
|
| 299 |
+
"grad_norm": 0.1736782193183899,
|
| 300 |
+
"learning_rate": 5.275155131394825e-06,
|
| 301 |
+
"loss": 0.5047,
|
| 302 |
+
"mean_token_accuracy": 0.854168564081192,
|
| 303 |
+
"num_tokens": 1017950.0,
|
| 304 |
+
"step": 250
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.077837837837838,
|
| 308 |
+
"eval_entropy": 0.5199673544706368,
|
| 309 |
+
"eval_loss": 0.5092993974685669,
|
| 310 |
+
"eval_mean_token_accuracy": 0.855072995511497,
|
| 311 |
+
"eval_num_tokens": 1017950.0,
|
| 312 |
+
"eval_runtime": 40.6579,
|
| 313 |
+
"eval_samples_per_second": 4.034,
|
| 314 |
+
"eval_steps_per_second": 4.034,
|
| 315 |
+
"step": 250
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"entropy": 0.4647901371121407,
|
| 319 |
+
"epoch": 1.1210810810810812,
|
| 320 |
+
"grad_norm": 0.18243983387947083,
|
| 321 |
+
"learning_rate": 4.951419434695115e-06,
|
| 322 |
+
"loss": 0.4514,
|
| 323 |
+
"mean_token_accuracy": 0.8703323498368263,
|
| 324 |
+
"num_tokens": 1058910.0,
|
| 325 |
+
"step": 260
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"entropy": 0.4522815000265837,
|
| 329 |
+
"epoch": 1.1643243243243244,
|
| 330 |
+
"grad_norm": 0.20347489416599274,
|
| 331 |
+
"learning_rate": 4.627887501769231e-06,
|
| 332 |
+
"loss": 0.4454,
|
| 333 |
+
"mean_token_accuracy": 0.8707166478037834,
|
| 334 |
+
"num_tokens": 1099604.0,
|
| 335 |
+
"step": 270
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"entropy": 0.4927990548312664,
|
| 339 |
+
"epoch": 1.2075675675675677,
|
| 340 |
+
"grad_norm": 0.18338286876678467,
|
| 341 |
+
"learning_rate": 4.3059163379791676e-06,
|
| 342 |
+
"loss": 0.491,
|
| 343 |
+
"mean_token_accuracy": 0.8596193596720696,
|
| 344 |
+
"num_tokens": 1140464.0,
|
| 345 |
+
"step": 280
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"entropy": 0.4583901420235634,
|
| 349 |
+
"epoch": 1.250810810810811,
|
| 350 |
+
"grad_norm": 0.18116827309131622,
|
| 351 |
+
"learning_rate": 3.98685640227875e-06,
|
| 352 |
+
"loss": 0.4446,
|
| 353 |
+
"mean_token_accuracy": 0.8701988190412522,
|
| 354 |
+
"num_tokens": 1181280.0,
|
| 355 |
+
"step": 290
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"entropy": 0.5115556083619595,
|
| 359 |
+
"epoch": 1.2940540540540542,
|
| 360 |
+
"grad_norm": 0.21102362871170044,
|
| 361 |
+
"learning_rate": 3.6720459429194743e-06,
|
| 362 |
+
"loss": 0.5013,
|
| 363 |
+
"mean_token_accuracy": 0.8555837154388428,
|
| 364 |
+
"num_tokens": 1221884.0,
|
| 365 |
+
"step": 300
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.2940540540540542,
|
| 369 |
+
"eval_entropy": 0.5066255062091641,
|
| 370 |
+
"eval_loss": 0.5047101974487305,
|
| 371 |
+
"eval_mean_token_accuracy": 0.856273435964817,
|
| 372 |
+
"eval_num_tokens": 1221884.0,
|
| 373 |
+
"eval_runtime": 40.5902,
|
| 374 |
+
"eval_samples_per_second": 4.04,
|
| 375 |
+
"eval_steps_per_second": 4.04,
|
| 376 |
+
"step": 300
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"entropy": 0.5146898411214351,
|
| 380 |
+
"epoch": 1.3372972972972974,
|
| 381 |
+
"grad_norm": 0.23965492844581604,
|
| 382 |
+
"learning_rate": 3.3628053843722674e-06,
|
| 383 |
+
"loss": 0.5126,
|
| 384 |
+
"mean_token_accuracy": 0.8517826452851296,
|
| 385 |
+
"num_tokens": 1262318.0,
|
| 386 |
+
"step": 310
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"entropy": 0.45455632023513315,
|
| 390 |
+
"epoch": 1.3805405405405407,
|
| 391 |
+
"grad_norm": 0.20733971893787384,
|
| 392 |
+
"learning_rate": 3.060431789008368e-06,
|
| 393 |
+
"loss": 0.4537,
|
| 394 |
+
"mean_token_accuracy": 0.87111437022686,
|
| 395 |
+
"num_tokens": 1303278.0,
|
| 396 |
+
"step": 320
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"entropy": 0.4839397206902504,
|
| 400 |
+
"epoch": 1.423783783783784,
|
| 401 |
+
"grad_norm": 0.20650845766067505,
|
| 402 |
+
"learning_rate": 2.7661934167689887e-06,
|
| 403 |
+
"loss": 0.4806,
|
| 404 |
+
"mean_token_accuracy": 0.8623866036534309,
|
| 405 |
+
"num_tokens": 1344200.0,
|
| 406 |
+
"step": 330
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"entropy": 0.5152107834815979,
|
| 410 |
+
"epoch": 1.4670270270270271,
|
| 411 |
+
"grad_norm": 0.1943357288837433,
|
| 412 |
+
"learning_rate": 2.4813244056423692e-06,
|
| 413 |
+
"loss": 0.5093,
|
| 414 |
+
"mean_token_accuracy": 0.8530791759490967,
|
| 415 |
+
"num_tokens": 1385160.0,
|
| 416 |
+
"step": 340
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"entropy": 0.5210597518831491,
|
| 420 |
+
"epoch": 1.5102702702702704,
|
| 421 |
+
"grad_norm": 0.23628048598766327,
|
| 422 |
+
"learning_rate": 2.207019595260154e-06,
|
| 423 |
+
"loss": 0.5187,
|
| 424 |
+
"mean_token_accuracy": 0.852712607383728,
|
| 425 |
+
"num_tokens": 1426120.0,
|
| 426 |
+
"step": 350
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 1.5102702702702704,
|
| 430 |
+
"eval_entropy": 0.5058667084611044,
|
| 431 |
+
"eval_loss": 0.5020793080329895,
|
| 432 |
+
"eval_mean_token_accuracy": 0.8568823802761916,
|
| 433 |
+
"eval_num_tokens": 1426120.0,
|
| 434 |
+
"eval_runtime": 40.5825,
|
| 435 |
+
"eval_samples_per_second": 4.041,
|
| 436 |
+
"eval_steps_per_second": 4.041,
|
| 437 |
+
"step": 350
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"entropy": 0.4884943708777428,
|
| 441 |
+
"epoch": 1.5535135135135136,
|
| 442 |
+
"grad_norm": 0.21701589226722717,
|
| 443 |
+
"learning_rate": 1.944429515324749e-06,
|
| 444 |
+
"loss": 0.4807,
|
| 445 |
+
"mean_token_accuracy": 0.8608564972877503,
|
| 446 |
+
"num_tokens": 1467012.0,
|
| 447 |
+
"step": 360
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"entropy": 0.4574359286576509,
|
| 451 |
+
"epoch": 1.5967567567567569,
|
| 452 |
+
"grad_norm": 0.1858481615781784,
|
| 453 |
+
"learning_rate": 1.6946555598879138e-06,
|
| 454 |
+
"loss": 0.4461,
|
| 455 |
+
"mean_token_accuracy": 0.8694037109613418,
|
| 456 |
+
"num_tokens": 1507972.0,
|
| 457 |
+
"step": 370
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"entropy": 0.49515259340405465,
|
| 461 |
+
"epoch": 1.6400000000000001,
|
| 462 |
+
"grad_norm": 0.1672651767730713,
|
| 463 |
+
"learning_rate": 1.4587453677213348e-06,
|
| 464 |
+
"loss": 0.4944,
|
| 465 |
+
"mean_token_accuracy": 0.8575757578015327,
|
| 466 |
+
"num_tokens": 1548932.0,
|
| 467 |
+
"step": 380
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"entropy": 0.48286144956946375,
|
| 471 |
+
"epoch": 1.6832432432432434,
|
| 472 |
+
"grad_norm": 0.1788497269153595,
|
| 473 |
+
"learning_rate": 1.2376884281555485e-06,
|
| 474 |
+
"loss": 0.4801,
|
| 475 |
+
"mean_token_accuracy": 0.8633186712861061,
|
| 476 |
+
"num_tokens": 1589892.0,
|
| 477 |
+
"step": 390
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"entropy": 0.42900044918060304,
|
| 481 |
+
"epoch": 1.7264864864864866,
|
| 482 |
+
"grad_norm": 0.18593338131904602,
|
| 483 |
+
"learning_rate": 1.0324119308178166e-06,
|
| 484 |
+
"loss": 0.4213,
|
| 485 |
+
"mean_token_accuracy": 0.8786656886339188,
|
| 486 |
+
"num_tokens": 1630852.0,
|
| 487 |
+
"step": 400
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.7264864864864866,
|
| 491 |
+
"eval_entropy": 0.5021736550985313,
|
| 492 |
+
"eval_loss": 0.5008729100227356,
|
| 493 |
+
"eval_mean_token_accuracy": 0.8570682758965144,
|
| 494 |
+
"eval_num_tokens": 1630852.0,
|
| 495 |
+
"eval_runtime": 40.6196,
|
| 496 |
+
"eval_samples_per_second": 4.037,
|
| 497 |
+
"eval_steps_per_second": 4.037,
|
| 498 |
+
"step": 400
|
| 499 |
+
}
|
| 500 |
+
],
|
| 501 |
+
"logging_steps": 10,
|
| 502 |
+
"max_steps": 500,
|
| 503 |
+
"num_input_tokens_seen": 0,
|
| 504 |
+
"num_train_epochs": 3,
|
| 505 |
+
"save_steps": 100,
|
| 506 |
+
"stateful_callbacks": {
|
| 507 |
+
"TrainerControl": {
|
| 508 |
+
"args": {
|
| 509 |
+
"should_epoch_stop": false,
|
| 510 |
+
"should_evaluate": false,
|
| 511 |
+
"should_log": false,
|
| 512 |
+
"should_save": true,
|
| 513 |
+
"should_training_stop": false
|
| 514 |
+
},
|
| 515 |
+
"attributes": {}
|
| 516 |
+
}
|
| 517 |
+
},
|
| 518 |
+
"total_flos": 7.492974219450778e+16,
|
| 519 |
+
"train_batch_size": 1,
|
| 520 |
+
"trial_name": null,
|
| 521 |
+
"trial_params": null
|
| 522 |
+
}
|
checkpoint-500/adapter_config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "/home/support/llm/Ministral-8B-Instruct-2410/",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"corda_config": null,
|
| 7 |
+
"eva_config": null,
|
| 8 |
+
"exclude_modules": null,
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layer_replication": null,
|
| 13 |
+
"layers_pattern": null,
|
| 14 |
+
"layers_to_transform": null,
|
| 15 |
+
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 16,
|
| 17 |
+
"lora_bias": false,
|
| 18 |
+
"lora_dropout": 0.05,
|
| 19 |
+
"megatron_config": null,
|
| 20 |
+
"megatron_core": "megatron.core",
|
| 21 |
+
"modules_to_save": null,
|
| 22 |
+
"peft_type": "LORA",
|
| 23 |
+
"qalora_group_size": 16,
|
| 24 |
+
"r": 64,
|
| 25 |
+
"rank_pattern": {},
|
| 26 |
+
"revision": null,
|
| 27 |
+
"target_modules": [
|
| 28 |
+
"up_proj",
|
| 29 |
+
"down_proj",
|
| 30 |
+
"k_proj",
|
| 31 |
+
"gate_proj",
|
| 32 |
+
"v_proj",
|
| 33 |
+
"q_proj",
|
| 34 |
+
"o_proj"
|
| 35 |
+
],
|
| 36 |
+
"target_parameters": null,
|
| 37 |
+
"task_type": "CAUSAL_LM",
|
| 38 |
+
"trainable_token_indices": null,
|
| 39 |
+
"use_dora": false,
|
| 40 |
+
"use_qalora": false,
|
| 41 |
+
"use_rslora": false
|
| 42 |
+
}
|
checkpoint-500/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-500/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7edbeaf20dd7f571b5dd1c54d9ace4f9b6299127cc7ba2afb14a6d51a4a79a4
|
| 3 |
+
size 17078136
|
checkpoint-500/tokenizer_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-500/trainer_state.json
ADDED
|
@@ -0,0 +1,644 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 500,
|
| 3 |
+
"best_metric": 0.500381350517273,
|
| 4 |
+
"best_model_checkpoint": "./python-code-cot-ministral-8B/checkpoint-500",
|
| 5 |
+
"epoch": 2.155675675675676,
|
| 6 |
+
"eval_steps": 50,
|
| 7 |
+
"global_step": 500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 0.8707441985607147,
|
| 14 |
+
"epoch": 0.043243243243243246,
|
| 15 |
+
"grad_norm": 0.2866508960723877,
|
| 16 |
+
"learning_rate": 6e-06,
|
| 17 |
+
"loss": 1.065,
|
| 18 |
+
"mean_token_accuracy": 0.7471851870417595,
|
| 19 |
+
"num_tokens": 40787.0,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 0.818158182501793,
|
| 24 |
+
"epoch": 0.08648648648648649,
|
| 25 |
+
"grad_norm": 0.2426566779613495,
|
| 26 |
+
"learning_rate": 9.998321768134101e-06,
|
| 27 |
+
"loss": 0.9899,
|
| 28 |
+
"mean_token_accuracy": 0.7604838714003563,
|
| 29 |
+
"num_tokens": 81747.0,
|
| 30 |
+
"step": 20
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 0.8590745508670807,
|
| 34 |
+
"epoch": 0.12972972972972974,
|
| 35 |
+
"grad_norm": 0.2603059709072113,
|
| 36 |
+
"learning_rate": 9.979454595377594e-06,
|
| 37 |
+
"loss": 0.9586,
|
| 38 |
+
"mean_token_accuracy": 0.7590420320630074,
|
| 39 |
+
"num_tokens": 122707.0,
|
| 40 |
+
"step": 30
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 0.7979068487882615,
|
| 44 |
+
"epoch": 0.17297297297297298,
|
| 45 |
+
"grad_norm": 0.2751409411430359,
|
| 46 |
+
"learning_rate": 9.93970185956522e-06,
|
| 47 |
+
"loss": 0.809,
|
| 48 |
+
"mean_token_accuracy": 0.7851744830608368,
|
| 49 |
+
"num_tokens": 163305.0,
|
| 50 |
+
"step": 40
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"entropy": 0.7995434030890465,
|
| 54 |
+
"epoch": 0.21621621621621623,
|
| 55 |
+
"grad_norm": 0.27979347109794617,
|
| 56 |
+
"learning_rate": 9.879230297486034e-06,
|
| 57 |
+
"loss": 0.7849,
|
| 58 |
+
"mean_token_accuracy": 0.791813300549984,
|
| 59 |
+
"num_tokens": 204265.0,
|
| 60 |
+
"step": 50
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.21621621621621623,
|
| 64 |
+
"eval_entropy": 0.7247167360491868,
|
| 65 |
+
"eval_loss": 0.7107394337654114,
|
| 66 |
+
"eval_mean_token_accuracy": 0.813098669052124,
|
| 67 |
+
"eval_num_tokens": 204265.0,
|
| 68 |
+
"eval_runtime": 40.5827,
|
| 69 |
+
"eval_samples_per_second": 4.041,
|
| 70 |
+
"eval_steps_per_second": 4.041,
|
| 71 |
+
"step": 50
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"entropy": 0.6114772140979767,
|
| 75 |
+
"epoch": 0.2594594594594595,
|
| 76 |
+
"grad_norm": 0.1919497400522232,
|
| 77 |
+
"learning_rate": 9.798293547886748e-06,
|
| 78 |
+
"loss": 0.6123,
|
| 79 |
+
"mean_token_accuracy": 0.8428526982665062,
|
| 80 |
+
"num_tokens": 245044.0,
|
| 81 |
+
"step": 60
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"entropy": 0.5336497366428375,
|
| 85 |
+
"epoch": 0.3027027027027027,
|
| 86 |
+
"grad_norm": 0.20059765875339508,
|
| 87 |
+
"learning_rate": 9.697231087622691e-06,
|
| 88 |
+
"loss": 0.5426,
|
| 89 |
+
"mean_token_accuracy": 0.8581829935312271,
|
| 90 |
+
"num_tokens": 285929.0,
|
| 91 |
+
"step": 70
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"entropy": 0.49029816724359987,
|
| 95 |
+
"epoch": 0.34594594594594597,
|
| 96 |
+
"grad_norm": 0.19071193039417267,
|
| 97 |
+
"learning_rate": 9.5764668077739e-06,
|
| 98 |
+
"loss": 0.5078,
|
| 99 |
+
"mean_token_accuracy": 0.8675507545471192,
|
| 100 |
+
"num_tokens": 326731.0,
|
| 101 |
+
"step": 80
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"entropy": 0.568328783661127,
|
| 105 |
+
"epoch": 0.3891891891891892,
|
| 106 |
+
"grad_norm": 0.1908392310142517,
|
| 107 |
+
"learning_rate": 9.436507235698613e-06,
|
| 108 |
+
"loss": 0.5825,
|
| 109 |
+
"mean_token_accuracy": 0.8428896099328995,
|
| 110 |
+
"num_tokens": 367252.0,
|
| 111 |
+
"step": 90
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"entropy": 0.5580322951078415,
|
| 115 |
+
"epoch": 0.43243243243243246,
|
| 116 |
+
"grad_norm": 0.1762734353542328,
|
| 117 |
+
"learning_rate": 9.277939410481507e-06,
|
| 118 |
+
"loss": 0.5536,
|
| 119 |
+
"mean_token_accuracy": 0.8481671541929245,
|
| 120 |
+
"num_tokens": 408212.0,
|
| 121 |
+
"step": 100
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.43243243243243246,
|
| 125 |
+
"eval_entropy": 0.5617244329939528,
|
| 126 |
+
"eval_loss": 0.5632628202438354,
|
| 127 |
+
"eval_mean_token_accuracy": 0.8465738710833759,
|
| 128 |
+
"eval_num_tokens": 408212.0,
|
| 129 |
+
"eval_runtime": 40.5739,
|
| 130 |
+
"eval_samples_per_second": 4.042,
|
| 131 |
+
"eval_steps_per_second": 4.042,
|
| 132 |
+
"step": 100
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 0.5392510816454887,
|
| 136 |
+
"epoch": 0.4756756756756757,
|
| 137 |
+
"grad_norm": 0.13735364377498627,
|
| 138 |
+
"learning_rate": 9.101428420687759e-06,
|
| 139 |
+
"loss": 0.5391,
|
| 140 |
+
"mean_token_accuracy": 0.8521016612648964,
|
| 141 |
+
"num_tokens": 449172.0,
|
| 142 |
+
"step": 110
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 0.6197883028537035,
|
| 146 |
+
"epoch": 0.518918918918919,
|
| 147 |
+
"grad_norm": 0.1790686547756195,
|
| 148 |
+
"learning_rate": 8.907714614750473e-06,
|
| 149 |
+
"loss": 0.6066,
|
| 150 |
+
"mean_token_accuracy": 0.837218964099884,
|
| 151 |
+
"num_tokens": 490132.0,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"entropy": 0.5101678855717182,
|
| 156 |
+
"epoch": 0.5621621621621622,
|
| 157 |
+
"grad_norm": 0.16557781398296356,
|
| 158 |
+
"learning_rate": 8.697610495692055e-06,
|
| 159 |
+
"loss": 0.4937,
|
| 160 |
+
"mean_token_accuracy": 0.861264032125473,
|
| 161 |
+
"num_tokens": 531046.0,
|
| 162 |
+
"step": 130
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"entropy": 0.5256278648972511,
|
| 166 |
+
"epoch": 0.6054054054054054,
|
| 167 |
+
"grad_norm": 0.15723901987075806,
|
| 168 |
+
"learning_rate": 8.471997313204183e-06,
|
| 169 |
+
"loss": 0.5128,
|
| 170 |
+
"mean_token_accuracy": 0.8562316685914994,
|
| 171 |
+
"num_tokens": 572006.0,
|
| 172 |
+
"step": 140
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"entropy": 0.5276982985436917,
|
| 176 |
+
"epoch": 0.6486486486486487,
|
| 177 |
+
"grad_norm": 0.14954417943954468,
|
| 178 |
+
"learning_rate": 8.231821367380335e-06,
|
| 179 |
+
"loss": 0.5148,
|
| 180 |
+
"mean_token_accuracy": 0.8543255120515824,
|
| 181 |
+
"num_tokens": 612966.0,
|
| 182 |
+
"step": 150
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.6486486486486487,
|
| 186 |
+
"eval_entropy": 0.5507546397970944,
|
| 187 |
+
"eval_loss": 0.5337628722190857,
|
| 188 |
+
"eval_mean_token_accuracy": 0.8515133745059734,
|
| 189 |
+
"eval_num_tokens": 612966.0,
|
| 190 |
+
"eval_runtime": 40.52,
|
| 191 |
+
"eval_samples_per_second": 4.047,
|
| 192 |
+
"eval_steps_per_second": 4.047,
|
| 193 |
+
"step": 150
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"entropy": 0.5230918094515801,
|
| 197 |
+
"epoch": 0.6918918918918919,
|
| 198 |
+
"grad_norm": 0.14227904379367828,
|
| 199 |
+
"learning_rate": 7.978090039604342e-06,
|
| 200 |
+
"loss": 0.5118,
|
| 201 |
+
"mean_token_accuracy": 0.8542912498116493,
|
| 202 |
+
"num_tokens": 653680.0,
|
| 203 |
+
"step": 160
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"entropy": 0.524362176656723,
|
| 207 |
+
"epoch": 0.7351351351351352,
|
| 208 |
+
"grad_norm": 0.16846755146980286,
|
| 209 |
+
"learning_rate": 7.711867567242769e-06,
|
| 210 |
+
"loss": 0.5104,
|
| 211 |
+
"mean_token_accuracy": 0.8555971384048462,
|
| 212 |
+
"num_tokens": 694478.0,
|
| 213 |
+
"step": 170
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"entropy": 0.49619704745709897,
|
| 217 |
+
"epoch": 0.7783783783783784,
|
| 218 |
+
"grad_norm": 0.20149804651737213,
|
| 219 |
+
"learning_rate": 7.434270579863549e-06,
|
| 220 |
+
"loss": 0.4826,
|
| 221 |
+
"mean_token_accuracy": 0.8604745492339134,
|
| 222 |
+
"num_tokens": 735351.0,
|
| 223 |
+
"step": 180
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"entropy": 0.5439279317855835,
|
| 227 |
+
"epoch": 0.8216216216216217,
|
| 228 |
+
"grad_norm": 0.22333768010139465,
|
| 229 |
+
"learning_rate": 7.14646341570353e-06,
|
| 230 |
+
"loss": 0.5411,
|
| 231 |
+
"mean_token_accuracy": 0.8482905030250549,
|
| 232 |
+
"num_tokens": 775651.0,
|
| 233 |
+
"step": 190
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"entropy": 0.4976503163576126,
|
| 237 |
+
"epoch": 0.8648648648648649,
|
| 238 |
+
"grad_norm": 0.16309203207492828,
|
| 239 |
+
"learning_rate": 6.849653238029261e-06,
|
| 240 |
+
"loss": 0.4994,
|
| 241 |
+
"mean_token_accuracy": 0.85667155534029,
|
| 242 |
+
"num_tokens": 816611.0,
|
| 243 |
+
"step": 200
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.8648648648648649,
|
| 247 |
+
"eval_entropy": 0.5280105926823325,
|
| 248 |
+
"eval_loss": 0.5184423327445984,
|
| 249 |
+
"eval_mean_token_accuracy": 0.8534702516910506,
|
| 250 |
+
"eval_num_tokens": 816611.0,
|
| 251 |
+
"eval_runtime": 40.6179,
|
| 252 |
+
"eval_samples_per_second": 4.038,
|
| 253 |
+
"eval_steps_per_second": 4.038,
|
| 254 |
+
"step": 200
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"entropy": 0.5327252022922039,
|
| 258 |
+
"epoch": 0.9081081081081082,
|
| 259 |
+
"grad_norm": 0.16586290299892426,
|
| 260 |
+
"learning_rate": 6.545084971874738e-06,
|
| 261 |
+
"loss": 0.5268,
|
| 262 |
+
"mean_token_accuracy": 0.8529814258217812,
|
| 263 |
+
"num_tokens": 857571.0,
|
| 264 |
+
"step": 210
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"entropy": 0.47181336022913456,
|
| 268 |
+
"epoch": 0.9513513513513514,
|
| 269 |
+
"grad_norm": 0.17959456145763397,
|
| 270 |
+
"learning_rate": 6.234036082393171e-06,
|
| 271 |
+
"loss": 0.4687,
|
| 272 |
+
"mean_token_accuracy": 0.8678423285484314,
|
| 273 |
+
"num_tokens": 898415.0,
|
| 274 |
+
"step": 220
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"entropy": 0.5040528390556573,
|
| 278 |
+
"epoch": 0.9945945945945946,
|
| 279 |
+
"grad_norm": 0.17449511587619781,
|
| 280 |
+
"learning_rate": 5.9178112167241805e-06,
|
| 281 |
+
"loss": 0.4999,
|
| 282 |
+
"mean_token_accuracy": 0.8588241055607796,
|
| 283 |
+
"num_tokens": 939275.0,
|
| 284 |
+
"step": 230
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"entropy": 0.4610033663543495,
|
| 288 |
+
"epoch": 1.0345945945945947,
|
| 289 |
+
"grad_norm": 0.15696631371974945,
|
| 290 |
+
"learning_rate": 5.597736731850295e-06,
|
| 291 |
+
"loss": 0.4834,
|
| 292 |
+
"mean_token_accuracy": 0.8710998361175125,
|
| 293 |
+
"num_tokens": 977163.0,
|
| 294 |
+
"step": 240
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"entropy": 0.5211307317018509,
|
| 298 |
+
"epoch": 1.077837837837838,
|
| 299 |
+
"grad_norm": 0.1736782193183899,
|
| 300 |
+
"learning_rate": 5.275155131394825e-06,
|
| 301 |
+
"loss": 0.5047,
|
| 302 |
+
"mean_token_accuracy": 0.854168564081192,
|
| 303 |
+
"num_tokens": 1017950.0,
|
| 304 |
+
"step": 250
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.077837837837838,
|
| 308 |
+
"eval_entropy": 0.5199673544706368,
|
| 309 |
+
"eval_loss": 0.5092993974685669,
|
| 310 |
+
"eval_mean_token_accuracy": 0.855072995511497,
|
| 311 |
+
"eval_num_tokens": 1017950.0,
|
| 312 |
+
"eval_runtime": 40.6579,
|
| 313 |
+
"eval_samples_per_second": 4.034,
|
| 314 |
+
"eval_steps_per_second": 4.034,
|
| 315 |
+
"step": 250
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"entropy": 0.4647901371121407,
|
| 319 |
+
"epoch": 1.1210810810810812,
|
| 320 |
+
"grad_norm": 0.18243983387947083,
|
| 321 |
+
"learning_rate": 4.951419434695115e-06,
|
| 322 |
+
"loss": 0.4514,
|
| 323 |
+
"mean_token_accuracy": 0.8703323498368263,
|
| 324 |
+
"num_tokens": 1058910.0,
|
| 325 |
+
"step": 260
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"entropy": 0.4522815000265837,
|
| 329 |
+
"epoch": 1.1643243243243244,
|
| 330 |
+
"grad_norm": 0.20347489416599274,
|
| 331 |
+
"learning_rate": 4.627887501769231e-06,
|
| 332 |
+
"loss": 0.4454,
|
| 333 |
+
"mean_token_accuracy": 0.8707166478037834,
|
| 334 |
+
"num_tokens": 1099604.0,
|
| 335 |
+
"step": 270
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"entropy": 0.4927990548312664,
|
| 339 |
+
"epoch": 1.2075675675675677,
|
| 340 |
+
"grad_norm": 0.18338286876678467,
|
| 341 |
+
"learning_rate": 4.3059163379791676e-06,
|
| 342 |
+
"loss": 0.491,
|
| 343 |
+
"mean_token_accuracy": 0.8596193596720696,
|
| 344 |
+
"num_tokens": 1140464.0,
|
| 345 |
+
"step": 280
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"entropy": 0.4583901420235634,
|
| 349 |
+
"epoch": 1.250810810810811,
|
| 350 |
+
"grad_norm": 0.18116827309131622,
|
| 351 |
+
"learning_rate": 3.98685640227875e-06,
|
| 352 |
+
"loss": 0.4446,
|
| 353 |
+
"mean_token_accuracy": 0.8701988190412522,
|
| 354 |
+
"num_tokens": 1181280.0,
|
| 355 |
+
"step": 290
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"entropy": 0.5115556083619595,
|
| 359 |
+
"epoch": 1.2940540540540542,
|
| 360 |
+
"grad_norm": 0.21102362871170044,
|
| 361 |
+
"learning_rate": 3.6720459429194743e-06,
|
| 362 |
+
"loss": 0.5013,
|
| 363 |
+
"mean_token_accuracy": 0.8555837154388428,
|
| 364 |
+
"num_tokens": 1221884.0,
|
| 365 |
+
"step": 300
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.2940540540540542,
|
| 369 |
+
"eval_entropy": 0.5066255062091641,
|
| 370 |
+
"eval_loss": 0.5047101974487305,
|
| 371 |
+
"eval_mean_token_accuracy": 0.856273435964817,
|
| 372 |
+
"eval_num_tokens": 1221884.0,
|
| 373 |
+
"eval_runtime": 40.5902,
|
| 374 |
+
"eval_samples_per_second": 4.04,
|
| 375 |
+
"eval_steps_per_second": 4.04,
|
| 376 |
+
"step": 300
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"entropy": 0.5146898411214351,
|
| 380 |
+
"epoch": 1.3372972972972974,
|
| 381 |
+
"grad_norm": 0.23965492844581604,
|
| 382 |
+
"learning_rate": 3.3628053843722674e-06,
|
| 383 |
+
"loss": 0.5126,
|
| 384 |
+
"mean_token_accuracy": 0.8517826452851296,
|
| 385 |
+
"num_tokens": 1262318.0,
|
| 386 |
+
"step": 310
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"entropy": 0.45455632023513315,
|
| 390 |
+
"epoch": 1.3805405405405407,
|
| 391 |
+
"grad_norm": 0.20733971893787384,
|
| 392 |
+
"learning_rate": 3.060431789008368e-06,
|
| 393 |
+
"loss": 0.4537,
|
| 394 |
+
"mean_token_accuracy": 0.87111437022686,
|
| 395 |
+
"num_tokens": 1303278.0,
|
| 396 |
+
"step": 320
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"entropy": 0.4839397206902504,
|
| 400 |
+
"epoch": 1.423783783783784,
|
| 401 |
+
"grad_norm": 0.20650845766067505,
|
| 402 |
+
"learning_rate": 2.7661934167689887e-06,
|
| 403 |
+
"loss": 0.4806,
|
| 404 |
+
"mean_token_accuracy": 0.8623866036534309,
|
| 405 |
+
"num_tokens": 1344200.0,
|
| 406 |
+
"step": 330
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"entropy": 0.5152107834815979,
|
| 410 |
+
"epoch": 1.4670270270270271,
|
| 411 |
+
"grad_norm": 0.1943357288837433,
|
| 412 |
+
"learning_rate": 2.4813244056423692e-06,
|
| 413 |
+
"loss": 0.5093,
|
| 414 |
+
"mean_token_accuracy": 0.8530791759490967,
|
| 415 |
+
"num_tokens": 1385160.0,
|
| 416 |
+
"step": 340
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"entropy": 0.5210597518831491,
|
| 420 |
+
"epoch": 1.5102702702702704,
|
| 421 |
+
"grad_norm": 0.23628048598766327,
|
| 422 |
+
"learning_rate": 2.207019595260154e-06,
|
| 423 |
+
"loss": 0.5187,
|
| 424 |
+
"mean_token_accuracy": 0.852712607383728,
|
| 425 |
+
"num_tokens": 1426120.0,
|
| 426 |
+
"step": 350
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 1.5102702702702704,
|
| 430 |
+
"eval_entropy": 0.5058667084611044,
|
| 431 |
+
"eval_loss": 0.5020793080329895,
|
| 432 |
+
"eval_mean_token_accuracy": 0.8568823802761916,
|
| 433 |
+
"eval_num_tokens": 1426120.0,
|
| 434 |
+
"eval_runtime": 40.5825,
|
| 435 |
+
"eval_samples_per_second": 4.041,
|
| 436 |
+
"eval_steps_per_second": 4.041,
|
| 437 |
+
"step": 350
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"entropy": 0.4884943708777428,
|
| 441 |
+
"epoch": 1.5535135135135136,
|
| 442 |
+
"grad_norm": 0.21701589226722717,
|
| 443 |
+
"learning_rate": 1.944429515324749e-06,
|
| 444 |
+
"loss": 0.4807,
|
| 445 |
+
"mean_token_accuracy": 0.8608564972877503,
|
| 446 |
+
"num_tokens": 1467012.0,
|
| 447 |
+
"step": 360
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"entropy": 0.4574359286576509,
|
| 451 |
+
"epoch": 1.5967567567567569,
|
| 452 |
+
"grad_norm": 0.1858481615781784,
|
| 453 |
+
"learning_rate": 1.6946555598879138e-06,
|
| 454 |
+
"loss": 0.4461,
|
| 455 |
+
"mean_token_accuracy": 0.8694037109613418,
|
| 456 |
+
"num_tokens": 1507972.0,
|
| 457 |
+
"step": 370
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"entropy": 0.49515259340405465,
|
| 461 |
+
"epoch": 1.6400000000000001,
|
| 462 |
+
"grad_norm": 0.1672651767730713,
|
| 463 |
+
"learning_rate": 1.4587453677213348e-06,
|
| 464 |
+
"loss": 0.4944,
|
| 465 |
+
"mean_token_accuracy": 0.8575757578015327,
|
| 466 |
+
"num_tokens": 1548932.0,
|
| 467 |
+
"step": 380
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"entropy": 0.48286144956946375,
|
| 471 |
+
"epoch": 1.6832432432432434,
|
| 472 |
+
"grad_norm": 0.1788497269153595,
|
| 473 |
+
"learning_rate": 1.2376884281555485e-06,
|
| 474 |
+
"loss": 0.4801,
|
| 475 |
+
"mean_token_accuracy": 0.8633186712861061,
|
| 476 |
+
"num_tokens": 1589892.0,
|
| 477 |
+
"step": 390
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"entropy": 0.42900044918060304,
|
| 481 |
+
"epoch": 1.7264864864864866,
|
| 482 |
+
"grad_norm": 0.18593338131904602,
|
| 483 |
+
"learning_rate": 1.0324119308178166e-06,
|
| 484 |
+
"loss": 0.4213,
|
| 485 |
+
"mean_token_accuracy": 0.8786656886339188,
|
| 486 |
+
"num_tokens": 1630852.0,
|
| 487 |
+
"step": 400
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.7264864864864866,
|
| 491 |
+
"eval_entropy": 0.5021736550985313,
|
| 492 |
+
"eval_loss": 0.5008729100227356,
|
| 493 |
+
"eval_mean_token_accuracy": 0.8570682758965144,
|
| 494 |
+
"eval_num_tokens": 1630852.0,
|
| 495 |
+
"eval_runtime": 40.6196,
|
| 496 |
+
"eval_samples_per_second": 4.037,
|
| 497 |
+
"eval_steps_per_second": 4.037,
|
| 498 |
+
"step": 400
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"entropy": 0.48913049399852754,
|
| 502 |
+
"epoch": 1.7697297297297299,
|
| 503 |
+
"grad_norm": 0.21752701699733734,
|
| 504 |
+
"learning_rate": 8.437768766765975e-07,
|
| 505 |
+
"loss": 0.4811,
|
| 506 |
+
"mean_token_accuracy": 0.8607771530747413,
|
| 507 |
+
"num_tokens": 1671621.0,
|
| 508 |
+
"step": 410
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"entropy": 0.464761159196496,
|
| 512 |
+
"epoch": 1.812972972972973,
|
| 513 |
+
"grad_norm": 0.17507386207580566,
|
| 514 |
+
"learning_rate": 6.725744667042778e-07,
|
| 515 |
+
"loss": 0.4635,
|
| 516 |
+
"mean_token_accuracy": 0.8650545433163643,
|
| 517 |
+
"num_tokens": 1712403.0,
|
| 518 |
+
"step": 420
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"entropy": 0.4877123299986124,
|
| 522 |
+
"epoch": 1.8562162162162164,
|
| 523 |
+
"grad_norm": 0.2440086156129837,
|
| 524 |
+
"learning_rate": 5.195227833053273e-07,
|
| 525 |
+
"loss": 0.4834,
|
| 526 |
+
"mean_token_accuracy": 0.859622411429882,
|
| 527 |
+
"num_tokens": 1753002.0,
|
| 528 |
+
"step": 430
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"entropy": 0.5305571794509888,
|
| 532 |
+
"epoch": 1.8994594594594596,
|
| 533 |
+
"grad_norm": 0.2179015576839447,
|
| 534 |
+
"learning_rate": 3.852637784291424e-07,
|
| 535 |
+
"loss": 0.5241,
|
| 536 |
+
"mean_token_accuracy": 0.8504694834351539,
|
| 537 |
+
"num_tokens": 1793720.0,
|
| 538 |
+
"step": 440
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"entropy": 0.45578556656837466,
|
| 542 |
+
"epoch": 1.9427027027027028,
|
| 543 |
+
"grad_norm": 0.19060668349266052,
|
| 544 |
+
"learning_rate": 2.7036058100047723e-07,
|
| 545 |
+
"loss": 0.4482,
|
| 546 |
+
"mean_token_accuracy": 0.8694037184119224,
|
| 547 |
+
"num_tokens": 1834680.0,
|
| 548 |
+
"step": 450
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.9427027027027028,
|
| 552 |
+
"eval_entropy": 0.5042242559536201,
|
| 553 |
+
"eval_loss": 0.5004777312278748,
|
| 554 |
+
"eval_mean_token_accuracy": 0.857193261384964,
|
| 555 |
+
"eval_num_tokens": 1834680.0,
|
| 556 |
+
"eval_runtime": 40.4789,
|
| 557 |
+
"eval_samples_per_second": 4.051,
|
| 558 |
+
"eval_steps_per_second": 4.051,
|
| 559 |
+
"step": 450
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"entropy": 0.5145775370299817,
|
| 563 |
+
"epoch": 1.985945945945946,
|
| 564 |
+
"grad_norm": 0.21297377347946167,
|
| 565 |
+
"learning_rate": 1.7529513496103322e-07,
|
| 566 |
+
"loss": 0.5102,
|
| 567 |
+
"mean_token_accuracy": 0.8554473593831062,
|
| 568 |
+
"num_tokens": 1875478.0,
|
| 569 |
+
"step": 460
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"entropy": 0.44477229424425074,
|
| 573 |
+
"epoch": 2.025945945945946,
|
| 574 |
+
"grad_norm": 0.1841362714767456,
|
| 575 |
+
"learning_rate": 1.004661778290783e-07,
|
| 576 |
+
"loss": 0.4269,
|
| 577 |
+
"mean_token_accuracy": 0.8740781545639038,
|
| 578 |
+
"num_tokens": 1913266.0,
|
| 579 |
+
"step": 470
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"entropy": 0.46048690713942053,
|
| 583 |
+
"epoch": 2.0691891891891894,
|
| 584 |
+
"grad_norm": 0.18584655225276947,
|
| 585 |
+
"learning_rate": 4.618756825572612e-08,
|
| 586 |
+
"loss": 0.4539,
|
| 587 |
+
"mean_token_accuracy": 0.8682526469230651,
|
| 588 |
+
"num_tokens": 1954188.0,
|
| 589 |
+
"step": 480
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"entropy": 0.42898516096174716,
|
| 593 |
+
"epoch": 2.1124324324324326,
|
| 594 |
+
"grad_norm": 0.17722217738628387,
|
| 595 |
+
"learning_rate": 1.268696959267679e-08,
|
| 596 |
+
"loss": 0.4209,
|
| 597 |
+
"mean_token_accuracy": 0.8789589449763298,
|
| 598 |
+
"num_tokens": 1995148.0,
|
| 599 |
+
"step": 490
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"entropy": 0.4906971834599972,
|
| 603 |
+
"epoch": 2.155675675675676,
|
| 604 |
+
"grad_norm": 0.22306711971759796,
|
| 605 |
+
"learning_rate": 1.0489499300603279e-10,
|
| 606 |
+
"loss": 0.4877,
|
| 607 |
+
"mean_token_accuracy": 0.8584246680140495,
|
| 608 |
+
"num_tokens": 2035965.0,
|
| 609 |
+
"step": 500
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 2.155675675675676,
|
| 613 |
+
"eval_entropy": 0.5036718782673522,
|
| 614 |
+
"eval_loss": 0.500381350517273,
|
| 615 |
+
"eval_mean_token_accuracy": 0.8574583268747097,
|
| 616 |
+
"eval_num_tokens": 2035965.0,
|
| 617 |
+
"eval_runtime": 40.5814,
|
| 618 |
+
"eval_samples_per_second": 4.041,
|
| 619 |
+
"eval_steps_per_second": 4.041,
|
| 620 |
+
"step": 500
|
| 621 |
+
}
|
| 622 |
+
],
|
| 623 |
+
"logging_steps": 10,
|
| 624 |
+
"max_steps": 500,
|
| 625 |
+
"num_input_tokens_seen": 0,
|
| 626 |
+
"num_train_epochs": 3,
|
| 627 |
+
"save_steps": 100,
|
| 628 |
+
"stateful_callbacks": {
|
| 629 |
+
"TrainerControl": {
|
| 630 |
+
"args": {
|
| 631 |
+
"should_epoch_stop": false,
|
| 632 |
+
"should_evaluate": false,
|
| 633 |
+
"should_log": false,
|
| 634 |
+
"should_save": true,
|
| 635 |
+
"should_training_stop": true
|
| 636 |
+
},
|
| 637 |
+
"attributes": {}
|
| 638 |
+
}
|
| 639 |
+
},
|
| 640 |
+
"total_flos": 9.354272034926592e+16,
|
| 641 |
+
"train_batch_size": 1,
|
| 642 |
+
"trial_name": null,
|
| 643 |
+
"trial_params": null
|
| 644 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7edbeaf20dd7f571b5dd1c54d9ace4f9b6299127cc7ba2afb14a6d51a4a79a4
|
| 3 |
+
size 17078136
|
tokenizer_config.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|