Delete checkpoint-5000
Browse files- checkpoint-5000/config.json +0 -26
- checkpoint-5000/generation_config.json +0 -10
- checkpoint-5000/optimizer.pt +0 -3
- checkpoint-5000/pytorch_model.bin +0 -3
- checkpoint-5000/rng_state.pth +0 -3
- checkpoint-5000/scheduler.pt +0 -3
- checkpoint-5000/special_tokens_map.json +0 -23
- checkpoint-5000/tokenizer.json +0 -0
- checkpoint-5000/tokenizer.model +0 -3
- checkpoint-5000/tokenizer_config.json +0 -34
- checkpoint-5000/trainer_state.json +0 -619
- checkpoint-5000/training_args.bin +0 -3
checkpoint-5000/config.json
DELETED
|
@@ -1,26 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"_name_or_path": "./core2",
|
| 3 |
-
"architectures": [
|
| 4 |
-
"LlamaForCausalLM"
|
| 5 |
-
],
|
| 6 |
-
"bos_token_id": 1,
|
| 7 |
-
"eos_token_id": 2,
|
| 8 |
-
"hidden_act": "silu",
|
| 9 |
-
"hidden_size": 4096,
|
| 10 |
-
"initializer_range": 0.02,
|
| 11 |
-
"intermediate_size": 11008,
|
| 12 |
-
"max_position_embeddings": 4096,
|
| 13 |
-
"model_type": "llama",
|
| 14 |
-
"num_attention_heads": 32,
|
| 15 |
-
"num_hidden_layers": 1,
|
| 16 |
-
"num_key_value_heads": 32,
|
| 17 |
-
"pretraining_tp": 1,
|
| 18 |
-
"rms_norm_eps": 1e-06,
|
| 19 |
-
"rope_scaling": null,
|
| 20 |
-
"rope_theta": 10000.0,
|
| 21 |
-
"tie_word_embeddings": false,
|
| 22 |
-
"torch_dtype": "bfloat16",
|
| 23 |
-
"transformers_version": "4.34.0.dev0",
|
| 24 |
-
"use_cache": true,
|
| 25 |
-
"vocab_size": 32000
|
| 26 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/generation_config.json
DELETED
|
@@ -1,10 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"bos_token_id": 1,
|
| 3 |
-
"do_sample": true,
|
| 4 |
-
"eos_token_id": 2,
|
| 5 |
-
"max_length": 4096,
|
| 6 |
-
"pad_token_id": 0,
|
| 7 |
-
"temperature": 0.6,
|
| 8 |
-
"top_p": 0.9,
|
| 9 |
-
"transformers_version": "4.34.0.dev0"
|
| 10 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/optimizer.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:704531a229cf5dc6fe0f5961f4f2867a29e55bf4fddef7b8c204b1054dbd40ab
|
| 3 |
-
size 1858136133
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/pytorch_model.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:8cfe8c3937d90ccd5a9a795bca91d29acb89c21869e7bfe0b28a6999060659f0
|
| 3 |
-
size 929067029
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/rng_state.pth
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:b2278a87cdf86c3f9219223c847f6b27f6b7f15b8226b617f38936e8ff2cbcde
|
| 3 |
-
size 14575
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/scheduler.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:d5192a1ee8b34f588f041e4906a7eb6bf85582d9fd03cfb971cdcda7ba896f86
|
| 3 |
-
size 627
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/special_tokens_map.json
DELETED
|
@@ -1,23 +0,0 @@
|
|
| 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 |
-
"unk_token": {
|
| 17 |
-
"content": "<unk>",
|
| 18 |
-
"lstrip": false,
|
| 19 |
-
"normalized": false,
|
| 20 |
-
"rstrip": false,
|
| 21 |
-
"single_word": false
|
| 22 |
-
}
|
| 23 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/tokenizer.json
DELETED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-5000/tokenizer.model
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
-
size 499723
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/tokenizer_config.json
DELETED
|
@@ -1,34 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"bos_token": {
|
| 3 |
-
"__type": "AddedToken",
|
| 4 |
-
"content": "<s>",
|
| 5 |
-
"lstrip": false,
|
| 6 |
-
"normalized": false,
|
| 7 |
-
"rstrip": false,
|
| 8 |
-
"single_word": false
|
| 9 |
-
},
|
| 10 |
-
"clean_up_tokenization_spaces": false,
|
| 11 |
-
"eos_token": {
|
| 12 |
-
"__type": "AddedToken",
|
| 13 |
-
"content": "</s>",
|
| 14 |
-
"lstrip": false,
|
| 15 |
-
"normalized": false,
|
| 16 |
-
"rstrip": false,
|
| 17 |
-
"single_word": false
|
| 18 |
-
},
|
| 19 |
-
"legacy": false,
|
| 20 |
-
"model_max_length": 1000000000000000019884624838656,
|
| 21 |
-
"pad_token": null,
|
| 22 |
-
"padding_side": "right",
|
| 23 |
-
"sp_model_kwargs": {},
|
| 24 |
-
"tokenizer_class": "LlamaTokenizer",
|
| 25 |
-
"unk_token": {
|
| 26 |
-
"__type": "AddedToken",
|
| 27 |
-
"content": "<unk>",
|
| 28 |
-
"lstrip": false,
|
| 29 |
-
"normalized": false,
|
| 30 |
-
"rstrip": false,
|
| 31 |
-
"single_word": false
|
| 32 |
-
},
|
| 33 |
-
"use_default_system_prompt": true
|
| 34 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/trainer_state.json
DELETED
|
@@ -1,619 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_metric": null,
|
| 3 |
-
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.22732826770176803,
|
| 5 |
-
"eval_steps": 500,
|
| 6 |
-
"global_step": 5000,
|
| 7 |
-
"is_hyper_param_search": false,
|
| 8 |
-
"is_local_process_zero": true,
|
| 9 |
-
"is_world_process_zero": true,
|
| 10 |
-
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.0,
|
| 13 |
-
"learning_rate": 9.977266527234701e-05,
|
| 14 |
-
"loss": 2.6296,
|
| 15 |
-
"step": 50
|
| 16 |
-
},
|
| 17 |
-
{
|
| 18 |
-
"epoch": 0.0,
|
| 19 |
-
"learning_rate": 9.954533054469402e-05,
|
| 20 |
-
"loss": 2.6156,
|
| 21 |
-
"step": 100
|
| 22 |
-
},
|
| 23 |
-
{
|
| 24 |
-
"epoch": 0.01,
|
| 25 |
-
"learning_rate": 9.931799581704102e-05,
|
| 26 |
-
"loss": 2.6035,
|
| 27 |
-
"step": 150
|
| 28 |
-
},
|
| 29 |
-
{
|
| 30 |
-
"epoch": 0.01,
|
| 31 |
-
"learning_rate": 9.909066108938801e-05,
|
| 32 |
-
"loss": 2.5399,
|
| 33 |
-
"step": 200
|
| 34 |
-
},
|
| 35 |
-
{
|
| 36 |
-
"epoch": 0.01,
|
| 37 |
-
"learning_rate": 9.886332636173502e-05,
|
| 38 |
-
"loss": 2.5857,
|
| 39 |
-
"step": 250
|
| 40 |
-
},
|
| 41 |
-
{
|
| 42 |
-
"epoch": 0.01,
|
| 43 |
-
"learning_rate": 9.863599163408202e-05,
|
| 44 |
-
"loss": 2.6078,
|
| 45 |
-
"step": 300
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.02,
|
| 49 |
-
"learning_rate": 9.840865690642903e-05,
|
| 50 |
-
"loss": 2.5931,
|
| 51 |
-
"step": 350
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.02,
|
| 55 |
-
"learning_rate": 9.818132217877604e-05,
|
| 56 |
-
"loss": 2.5919,
|
| 57 |
-
"step": 400
|
| 58 |
-
},
|
| 59 |
-
{
|
| 60 |
-
"epoch": 0.02,
|
| 61 |
-
"learning_rate": 9.795398745112304e-05,
|
| 62 |
-
"loss": 2.59,
|
| 63 |
-
"step": 450
|
| 64 |
-
},
|
| 65 |
-
{
|
| 66 |
-
"epoch": 0.02,
|
| 67 |
-
"learning_rate": 9.772665272347005e-05,
|
| 68 |
-
"loss": 2.605,
|
| 69 |
-
"step": 500
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.03,
|
| 73 |
-
"learning_rate": 9.749931799581704e-05,
|
| 74 |
-
"loss": 2.6026,
|
| 75 |
-
"step": 550
|
| 76 |
-
},
|
| 77 |
-
{
|
| 78 |
-
"epoch": 0.03,
|
| 79 |
-
"learning_rate": 9.727198326816404e-05,
|
| 80 |
-
"loss": 2.5839,
|
| 81 |
-
"step": 600
|
| 82 |
-
},
|
| 83 |
-
{
|
| 84 |
-
"epoch": 0.03,
|
| 85 |
-
"learning_rate": 9.704464854051105e-05,
|
| 86 |
-
"loss": 2.5862,
|
| 87 |
-
"step": 650
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.03,
|
| 91 |
-
"learning_rate": 9.681731381285806e-05,
|
| 92 |
-
"loss": 2.609,
|
| 93 |
-
"step": 700
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.03,
|
| 97 |
-
"learning_rate": 9.658997908520506e-05,
|
| 98 |
-
"loss": 2.5759,
|
| 99 |
-
"step": 750
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 0.04,
|
| 103 |
-
"learning_rate": 9.636264435755207e-05,
|
| 104 |
-
"loss": 2.6046,
|
| 105 |
-
"step": 800
|
| 106 |
-
},
|
| 107 |
-
{
|
| 108 |
-
"epoch": 0.04,
|
| 109 |
-
"learning_rate": 9.613530962989907e-05,
|
| 110 |
-
"loss": 2.5811,
|
| 111 |
-
"step": 850
|
| 112 |
-
},
|
| 113 |
-
{
|
| 114 |
-
"epoch": 0.04,
|
| 115 |
-
"learning_rate": 9.590797490224606e-05,
|
| 116 |
-
"loss": 2.5797,
|
| 117 |
-
"step": 900
|
| 118 |
-
},
|
| 119 |
-
{
|
| 120 |
-
"epoch": 0.04,
|
| 121 |
-
"learning_rate": 9.568064017459307e-05,
|
| 122 |
-
"loss": 2.5867,
|
| 123 |
-
"step": 950
|
| 124 |
-
},
|
| 125 |
-
{
|
| 126 |
-
"epoch": 0.05,
|
| 127 |
-
"learning_rate": 9.545330544694008e-05,
|
| 128 |
-
"loss": 2.5927,
|
| 129 |
-
"step": 1000
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.05,
|
| 133 |
-
"learning_rate": 9.522597071928708e-05,
|
| 134 |
-
"loss": 2.568,
|
| 135 |
-
"step": 1050
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.05,
|
| 139 |
-
"learning_rate": 9.499863599163409e-05,
|
| 140 |
-
"loss": 2.6024,
|
| 141 |
-
"step": 1100
|
| 142 |
-
},
|
| 143 |
-
{
|
| 144 |
-
"epoch": 0.05,
|
| 145 |
-
"learning_rate": 9.477130126398109e-05,
|
| 146 |
-
"loss": 2.5936,
|
| 147 |
-
"step": 1150
|
| 148 |
-
},
|
| 149 |
-
{
|
| 150 |
-
"epoch": 0.05,
|
| 151 |
-
"learning_rate": 9.45439665363281e-05,
|
| 152 |
-
"loss": 2.605,
|
| 153 |
-
"step": 1200
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 0.06,
|
| 157 |
-
"learning_rate": 9.431663180867509e-05,
|
| 158 |
-
"loss": 2.5775,
|
| 159 |
-
"step": 1250
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 0.06,
|
| 163 |
-
"learning_rate": 9.40892970810221e-05,
|
| 164 |
-
"loss": 2.5752,
|
| 165 |
-
"step": 1300
|
| 166 |
-
},
|
| 167 |
-
{
|
| 168 |
-
"epoch": 0.06,
|
| 169 |
-
"learning_rate": 9.38619623533691e-05,
|
| 170 |
-
"loss": 2.5679,
|
| 171 |
-
"step": 1350
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.06,
|
| 175 |
-
"learning_rate": 9.36346276257161e-05,
|
| 176 |
-
"loss": 2.5856,
|
| 177 |
-
"step": 1400
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.07,
|
| 181 |
-
"learning_rate": 9.340729289806311e-05,
|
| 182 |
-
"loss": 2.5787,
|
| 183 |
-
"step": 1450
|
| 184 |
-
},
|
| 185 |
-
{
|
| 186 |
-
"epoch": 0.07,
|
| 187 |
-
"learning_rate": 9.317995817041012e-05,
|
| 188 |
-
"loss": 2.5875,
|
| 189 |
-
"step": 1500
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 0.07,
|
| 193 |
-
"learning_rate": 9.295262344275712e-05,
|
| 194 |
-
"loss": 2.5631,
|
| 195 |
-
"step": 1550
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 0.07,
|
| 199 |
-
"learning_rate": 9.272528871510412e-05,
|
| 200 |
-
"loss": 2.583,
|
| 201 |
-
"step": 1600
|
| 202 |
-
},
|
| 203 |
-
{
|
| 204 |
-
"epoch": 0.08,
|
| 205 |
-
"learning_rate": 9.249795398745112e-05,
|
| 206 |
-
"loss": 2.5609,
|
| 207 |
-
"step": 1650
|
| 208 |
-
},
|
| 209 |
-
{
|
| 210 |
-
"epoch": 0.08,
|
| 211 |
-
"learning_rate": 9.227061925979813e-05,
|
| 212 |
-
"loss": 2.587,
|
| 213 |
-
"step": 1700
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.08,
|
| 217 |
-
"learning_rate": 9.204328453214513e-05,
|
| 218 |
-
"loss": 2.5555,
|
| 219 |
-
"step": 1750
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.08,
|
| 223 |
-
"learning_rate": 9.181594980449214e-05,
|
| 224 |
-
"loss": 2.5488,
|
| 225 |
-
"step": 1800
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 0.08,
|
| 229 |
-
"learning_rate": 9.158861507683914e-05,
|
| 230 |
-
"loss": 2.5554,
|
| 231 |
-
"step": 1850
|
| 232 |
-
},
|
| 233 |
-
{
|
| 234 |
-
"epoch": 0.09,
|
| 235 |
-
"learning_rate": 9.136128034918615e-05,
|
| 236 |
-
"loss": 2.5408,
|
| 237 |
-
"step": 1900
|
| 238 |
-
},
|
| 239 |
-
{
|
| 240 |
-
"epoch": 0.09,
|
| 241 |
-
"learning_rate": 9.113394562153314e-05,
|
| 242 |
-
"loss": 2.582,
|
| 243 |
-
"step": 1950
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"epoch": 0.09,
|
| 247 |
-
"learning_rate": 9.090661089388015e-05,
|
| 248 |
-
"loss": 2.5533,
|
| 249 |
-
"step": 2000
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.09,
|
| 253 |
-
"learning_rate": 9.067927616622715e-05,
|
| 254 |
-
"loss": 2.5432,
|
| 255 |
-
"step": 2050
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.1,
|
| 259 |
-
"learning_rate": 9.045194143857416e-05,
|
| 260 |
-
"loss": 2.5867,
|
| 261 |
-
"step": 2100
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.1,
|
| 265 |
-
"learning_rate": 9.022460671092116e-05,
|
| 266 |
-
"loss": 2.5343,
|
| 267 |
-
"step": 2150
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 0.1,
|
| 271 |
-
"learning_rate": 8.999727198326817e-05,
|
| 272 |
-
"loss": 2.585,
|
| 273 |
-
"step": 2200
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 0.1,
|
| 277 |
-
"learning_rate": 8.976993725561517e-05,
|
| 278 |
-
"loss": 2.5679,
|
| 279 |
-
"step": 2250
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 0.1,
|
| 283 |
-
"learning_rate": 8.954260252796217e-05,
|
| 284 |
-
"loss": 2.5515,
|
| 285 |
-
"step": 2300
|
| 286 |
-
},
|
| 287 |
-
{
|
| 288 |
-
"epoch": 0.11,
|
| 289 |
-
"learning_rate": 8.931526780030917e-05,
|
| 290 |
-
"loss": 2.5713,
|
| 291 |
-
"step": 2350
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.11,
|
| 295 |
-
"learning_rate": 8.908793307265618e-05,
|
| 296 |
-
"loss": 2.5587,
|
| 297 |
-
"step": 2400
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.11,
|
| 301 |
-
"learning_rate": 8.886059834500318e-05,
|
| 302 |
-
"loss": 2.5774,
|
| 303 |
-
"step": 2450
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.11,
|
| 307 |
-
"learning_rate": 8.863326361735019e-05,
|
| 308 |
-
"loss": 2.551,
|
| 309 |
-
"step": 2500
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 0.12,
|
| 313 |
-
"learning_rate": 8.84059288896972e-05,
|
| 314 |
-
"loss": 2.5685,
|
| 315 |
-
"step": 2550
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 0.12,
|
| 319 |
-
"learning_rate": 8.81785941620442e-05,
|
| 320 |
-
"loss": 2.5707,
|
| 321 |
-
"step": 2600
|
| 322 |
-
},
|
| 323 |
-
{
|
| 324 |
-
"epoch": 0.12,
|
| 325 |
-
"learning_rate": 8.795125943439119e-05,
|
| 326 |
-
"loss": 2.568,
|
| 327 |
-
"step": 2650
|
| 328 |
-
},
|
| 329 |
-
{
|
| 330 |
-
"epoch": 0.12,
|
| 331 |
-
"learning_rate": 8.77239247067382e-05,
|
| 332 |
-
"loss": 2.5536,
|
| 333 |
-
"step": 2700
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"epoch": 0.13,
|
| 337 |
-
"learning_rate": 8.74965899790852e-05,
|
| 338 |
-
"loss": 2.5406,
|
| 339 |
-
"step": 2750
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.13,
|
| 343 |
-
"learning_rate": 8.726925525143221e-05,
|
| 344 |
-
"loss": 2.5572,
|
| 345 |
-
"step": 2800
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 0.13,
|
| 349 |
-
"learning_rate": 8.704192052377921e-05,
|
| 350 |
-
"loss": 2.5749,
|
| 351 |
-
"step": 2850
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"epoch": 0.13,
|
| 355 |
-
"learning_rate": 8.681458579612622e-05,
|
| 356 |
-
"loss": 2.5607,
|
| 357 |
-
"step": 2900
|
| 358 |
-
},
|
| 359 |
-
{
|
| 360 |
-
"epoch": 0.13,
|
| 361 |
-
"learning_rate": 8.658725106847322e-05,
|
| 362 |
-
"loss": 2.5612,
|
| 363 |
-
"step": 2950
|
| 364 |
-
},
|
| 365 |
-
{
|
| 366 |
-
"epoch": 0.14,
|
| 367 |
-
"learning_rate": 8.635991634082023e-05,
|
| 368 |
-
"loss": 2.5626,
|
| 369 |
-
"step": 3000
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 0.14,
|
| 373 |
-
"learning_rate": 8.613258161316724e-05,
|
| 374 |
-
"loss": 2.5677,
|
| 375 |
-
"step": 3050
|
| 376 |
-
},
|
| 377 |
-
{
|
| 378 |
-
"epoch": 0.14,
|
| 379 |
-
"learning_rate": 8.590524688551423e-05,
|
| 380 |
-
"loss": 2.5072,
|
| 381 |
-
"step": 3100
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 0.14,
|
| 385 |
-
"learning_rate": 8.567791215786123e-05,
|
| 386 |
-
"loss": 2.562,
|
| 387 |
-
"step": 3150
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 0.15,
|
| 391 |
-
"learning_rate": 8.545057743020824e-05,
|
| 392 |
-
"loss": 2.5786,
|
| 393 |
-
"step": 3200
|
| 394 |
-
},
|
| 395 |
-
{
|
| 396 |
-
"epoch": 0.15,
|
| 397 |
-
"learning_rate": 8.522324270255524e-05,
|
| 398 |
-
"loss": 2.5388,
|
| 399 |
-
"step": 3250
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 0.15,
|
| 403 |
-
"learning_rate": 8.499590797490225e-05,
|
| 404 |
-
"loss": 2.538,
|
| 405 |
-
"step": 3300
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"epoch": 0.15,
|
| 409 |
-
"learning_rate": 8.476857324724926e-05,
|
| 410 |
-
"loss": 2.5448,
|
| 411 |
-
"step": 3350
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 0.15,
|
| 415 |
-
"learning_rate": 8.454123851959626e-05,
|
| 416 |
-
"loss": 2.5306,
|
| 417 |
-
"step": 3400
|
| 418 |
-
},
|
| 419 |
-
{
|
| 420 |
-
"epoch": 0.16,
|
| 421 |
-
"learning_rate": 8.431390379194327e-05,
|
| 422 |
-
"loss": 2.5647,
|
| 423 |
-
"step": 3450
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 0.16,
|
| 427 |
-
"learning_rate": 8.408656906429027e-05,
|
| 428 |
-
"loss": 2.5386,
|
| 429 |
-
"step": 3500
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.16,
|
| 433 |
-
"learning_rate": 8.385923433663728e-05,
|
| 434 |
-
"loss": 2.5376,
|
| 435 |
-
"step": 3550
|
| 436 |
-
},
|
| 437 |
-
{
|
| 438 |
-
"epoch": 0.16,
|
| 439 |
-
"learning_rate": 8.363189960898427e-05,
|
| 440 |
-
"loss": 2.535,
|
| 441 |
-
"step": 3600
|
| 442 |
-
},
|
| 443 |
-
{
|
| 444 |
-
"epoch": 0.17,
|
| 445 |
-
"learning_rate": 8.340456488133128e-05,
|
| 446 |
-
"loss": 2.5476,
|
| 447 |
-
"step": 3650
|
| 448 |
-
},
|
| 449 |
-
{
|
| 450 |
-
"epoch": 0.17,
|
| 451 |
-
"learning_rate": 8.317723015367828e-05,
|
| 452 |
-
"loss": 2.5462,
|
| 453 |
-
"step": 3700
|
| 454 |
-
},
|
| 455 |
-
{
|
| 456 |
-
"epoch": 0.17,
|
| 457 |
-
"learning_rate": 8.294989542602529e-05,
|
| 458 |
-
"loss": 2.5795,
|
| 459 |
-
"step": 3750
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 0.17,
|
| 463 |
-
"learning_rate": 8.272256069837229e-05,
|
| 464 |
-
"loss": 2.5425,
|
| 465 |
-
"step": 3800
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 0.18,
|
| 469 |
-
"learning_rate": 8.24952259707193e-05,
|
| 470 |
-
"loss": 2.5663,
|
| 471 |
-
"step": 3850
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.18,
|
| 475 |
-
"learning_rate": 8.22678912430663e-05,
|
| 476 |
-
"loss": 2.5376,
|
| 477 |
-
"step": 3900
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"epoch": 0.18,
|
| 481 |
-
"learning_rate": 8.204055651541331e-05,
|
| 482 |
-
"loss": 2.5695,
|
| 483 |
-
"step": 3950
|
| 484 |
-
},
|
| 485 |
-
{
|
| 486 |
-
"epoch": 0.18,
|
| 487 |
-
"learning_rate": 8.18132217877603e-05,
|
| 488 |
-
"loss": 2.5112,
|
| 489 |
-
"step": 4000
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 0.18,
|
| 493 |
-
"learning_rate": 8.15858870601073e-05,
|
| 494 |
-
"loss": 2.5255,
|
| 495 |
-
"step": 4050
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"epoch": 0.19,
|
| 499 |
-
"learning_rate": 8.135855233245431e-05,
|
| 500 |
-
"loss": 2.5448,
|
| 501 |
-
"step": 4100
|
| 502 |
-
},
|
| 503 |
-
{
|
| 504 |
-
"epoch": 0.19,
|
| 505 |
-
"learning_rate": 8.113121760480132e-05,
|
| 506 |
-
"loss": 2.5483,
|
| 507 |
-
"step": 4150
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.19,
|
| 511 |
-
"learning_rate": 8.090388287714832e-05,
|
| 512 |
-
"loss": 2.5319,
|
| 513 |
-
"step": 4200
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.19,
|
| 517 |
-
"learning_rate": 8.067654814949533e-05,
|
| 518 |
-
"loss": 2.5655,
|
| 519 |
-
"step": 4250
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 0.2,
|
| 523 |
-
"learning_rate": 8.044921342184233e-05,
|
| 524 |
-
"loss": 2.5399,
|
| 525 |
-
"step": 4300
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 0.2,
|
| 529 |
-
"learning_rate": 8.022187869418933e-05,
|
| 530 |
-
"loss": 2.5485,
|
| 531 |
-
"step": 4350
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"epoch": 0.2,
|
| 535 |
-
"learning_rate": 7.999454396653633e-05,
|
| 536 |
-
"loss": 2.5196,
|
| 537 |
-
"step": 4400
|
| 538 |
-
},
|
| 539 |
-
{
|
| 540 |
-
"epoch": 0.2,
|
| 541 |
-
"learning_rate": 7.976720923888334e-05,
|
| 542 |
-
"loss": 2.5554,
|
| 543 |
-
"step": 4450
|
| 544 |
-
},
|
| 545 |
-
{
|
| 546 |
-
"epoch": 0.2,
|
| 547 |
-
"learning_rate": 7.953987451123034e-05,
|
| 548 |
-
"loss": 2.583,
|
| 549 |
-
"step": 4500
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.21,
|
| 553 |
-
"learning_rate": 7.931253978357735e-05,
|
| 554 |
-
"loss": 2.5643,
|
| 555 |
-
"step": 4550
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.21,
|
| 559 |
-
"learning_rate": 7.908520505592435e-05,
|
| 560 |
-
"loss": 2.5345,
|
| 561 |
-
"step": 4600
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"epoch": 0.21,
|
| 565 |
-
"learning_rate": 7.885787032827136e-05,
|
| 566 |
-
"loss": 2.5393,
|
| 567 |
-
"step": 4650
|
| 568 |
-
},
|
| 569 |
-
{
|
| 570 |
-
"epoch": 0.21,
|
| 571 |
-
"learning_rate": 7.863053560061835e-05,
|
| 572 |
-
"loss": 2.5349,
|
| 573 |
-
"step": 4700
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 0.22,
|
| 577 |
-
"learning_rate": 7.840320087296536e-05,
|
| 578 |
-
"loss": 2.54,
|
| 579 |
-
"step": 4750
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 0.22,
|
| 583 |
-
"learning_rate": 7.817586614531236e-05,
|
| 584 |
-
"loss": 2.5526,
|
| 585 |
-
"step": 4800
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 0.22,
|
| 589 |
-
"learning_rate": 7.794853141765937e-05,
|
| 590 |
-
"loss": 2.5419,
|
| 591 |
-
"step": 4850
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.22,
|
| 595 |
-
"learning_rate": 7.772119669000637e-05,
|
| 596 |
-
"loss": 2.5122,
|
| 597 |
-
"step": 4900
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.23,
|
| 601 |
-
"learning_rate": 7.749386196235338e-05,
|
| 602 |
-
"loss": 2.5247,
|
| 603 |
-
"step": 4950
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"epoch": 0.23,
|
| 607 |
-
"learning_rate": 7.726652723470039e-05,
|
| 608 |
-
"loss": 2.5516,
|
| 609 |
-
"step": 5000
|
| 610 |
-
}
|
| 611 |
-
],
|
| 612 |
-
"logging_steps": 50,
|
| 613 |
-
"max_steps": 21994,
|
| 614 |
-
"num_train_epochs": 1,
|
| 615 |
-
"save_steps": 2500,
|
| 616 |
-
"total_flos": 8.195099590656e+16,
|
| 617 |
-
"trial_name": null,
|
| 618 |
-
"trial_params": null
|
| 619 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoint-5000/training_args.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:efdafc920f92f7fef444725b8239e547a249e5e3d18c711b75245d4473de9d3c
|
| 3 |
-
size 4027
|
|
|
|
|
|
|
|
|
|
|
|