Training in progress, epoch 30, checkpoint
Browse files- checkpoint-10586/adapter_config.json +1 -1
- checkpoint-10586/adapter_model.bin +3 -0
- checkpoint-10586/added_tokens.json +3 -3
- checkpoint-10586/optimizer.pt +2 -2
- checkpoint-10586/rng_state_0.pth +1 -1
- checkpoint-10586/rng_state_1.pth +1 -1
- checkpoint-10586/rng_state_2.pth +1 -1
- checkpoint-10586/rng_state_3.pth +1 -1
- checkpoint-10586/special_tokens_map.json +4 -12
- checkpoint-10586/tokenizer.json +0 -0
- checkpoint-10586/tokenizer.model +3 -0
- checkpoint-10586/tokenizer_config.json +17 -20
- checkpoint-10586/trainer_state.json +435 -605
- checkpoint-10586/training_args.bin +2 -2
checkpoint-10586/adapter_config.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"auto_mapping": null,
|
| 3 |
-
"base_model_name_or_path": "
|
| 4 |
"bias": "none",
|
| 5 |
"fan_in_fan_out": false,
|
| 6 |
"inference_mode": true,
|
|
|
|
| 1 |
{
|
| 2 |
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 4 |
"bias": "none",
|
| 5 |
"fan_in_fan_out": false,
|
| 6 |
"inference_mode": true,
|
checkpoint-10586/adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6aac168cd920468bddb42bb2368aab42f91d90ec16e35dee7690a83aefe35211
|
| 3 |
+
size 143269386
|
checkpoint-10586/added_tokens.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
| 5 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"</s>": 2,
|
| 3 |
+
"<s>": 1,
|
| 4 |
+
"<unk>": 0
|
| 5 |
}
|
checkpoint-10586/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:983294e6588d685c6f760499b2a04f12f27e7273899d805111643a4a89e8a18e
|
| 3 |
+
size 286585234
|
checkpoint-10586/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e9965b823eebb89e74a9378ec18bb9ff83ff90bc7aa2288168c3f30987e615a9
|
| 3 |
size 15024
|
checkpoint-10586/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cbc9cc69b42c443126417dff63223e125b406a2822c049b9a9f8ab3030888f0e
|
| 3 |
size 15024
|
checkpoint-10586/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3bdcad8f9c113ab9d4974eea8d653911f69a5f728ec76e206811cf5576941401
|
| 3 |
size 15024
|
checkpoint-10586/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:40cfa5d8dfe1f2d28cd80fd327ffe9dc7823c16966ac12ff514111a172a24f9a
|
| 3 |
size 15024
|
checkpoint-10586/special_tokens_map.json
CHANGED
|
@@ -1,14 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
"eos_token": {
|
| 7 |
-
"content": "<|endoftext|>",
|
| 8 |
-
"lstrip": false,
|
| 9 |
-
"normalized": false,
|
| 10 |
-
"rstrip": false,
|
| 11 |
-
"single_word": false
|
| 12 |
-
},
|
| 13 |
-
"pad_token": "<|endoftext|>"
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"eos_token": "</s>",
|
| 4 |
+
"pad_token": "</s>",
|
| 5 |
+
"unk_token": "<unk>"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
}
|
checkpoint-10586/tokenizer.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-10586/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
checkpoint-10586/tokenizer_config.json
CHANGED
|
@@ -1,24 +1,23 @@
|
|
| 1 |
{
|
| 2 |
-
"add_prefix_space": false,
|
| 3 |
"added_tokens_decoder": {
|
| 4 |
-
"
|
| 5 |
-
"content": "
|
| 6 |
"lstrip": false,
|
| 7 |
"normalized": false,
|
| 8 |
"rstrip": false,
|
| 9 |
"single_word": false,
|
| 10 |
"special": true
|
| 11 |
},
|
| 12 |
-
"
|
| 13 |
-
"content": "
|
| 14 |
"lstrip": false,
|
| 15 |
"normalized": false,
|
| 16 |
"rstrip": false,
|
| 17 |
"single_word": false,
|
| 18 |
"special": true
|
| 19 |
},
|
| 20 |
-
"
|
| 21 |
-
"content": "
|
| 22 |
"lstrip": false,
|
| 23 |
"normalized": false,
|
| 24 |
"rstrip": false,
|
|
@@ -26,18 +25,16 @@
|
|
| 26 |
"special": true
|
| 27 |
}
|
| 28 |
},
|
| 29 |
-
"additional_special_tokens": [
|
| 30 |
-
|
| 31 |
-
"<|im_end|>"
|
| 32 |
-
],
|
| 33 |
-
"bos_token": null,
|
| 34 |
-
"chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
|
| 35 |
"clean_up_tokenization_spaces": false,
|
| 36 |
-
"eos_token": "
|
| 37 |
-
"
|
| 38 |
-
"model_max_length":
|
| 39 |
-
"pad_token": "
|
| 40 |
-
"
|
| 41 |
-
"
|
| 42 |
-
"
|
|
|
|
|
|
|
| 43 |
}
|
|
|
|
| 1 |
{
|
|
|
|
| 2 |
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
"lstrip": false,
|
| 6 |
"normalized": false,
|
| 7 |
"rstrip": false,
|
| 8 |
"single_word": false,
|
| 9 |
"special": true
|
| 10 |
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<s>",
|
| 13 |
"lstrip": false,
|
| 14 |
"normalized": false,
|
| 15 |
"rstrip": false,
|
| 16 |
"single_word": false,
|
| 17 |
"special": true
|
| 18 |
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
"lstrip": false,
|
| 22 |
"normalized": false,
|
| 23 |
"rstrip": false,
|
|
|
|
| 25 |
"special": true
|
| 26 |
}
|
| 27 |
},
|
| 28 |
+
"additional_special_tokens": [],
|
| 29 |
+
"bos_token": "<s>",
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
"clean_up_tokenization_spaces": false,
|
| 31 |
+
"eos_token": "</s>",
|
| 32 |
+
"legacy": false,
|
| 33 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 34 |
+
"pad_token": "</s>",
|
| 35 |
+
"padding_side": "left",
|
| 36 |
+
"sp_model_kwargs": {},
|
| 37 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 38 |
+
"unk_token": "<unk>",
|
| 39 |
+
"use_default_system_prompt": true
|
| 40 |
}
|
checkpoint-10586/trainer_state.json
CHANGED
|
@@ -9,1256 +9,1086 @@
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm": 0.2411503791809082,
|
| 14 |
"learning_rate": 0.0001,
|
| 15 |
-
"loss":
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 0.22286373376846313,
|
| 21 |
"learning_rate": 0.0001,
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm": 0.24079065024852753,
|
| 28 |
"learning_rate": 0.0001,
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0
|
| 34 |
-
"eval_accuracy": 0.
|
| 35 |
-
"eval_loss": 3.
|
| 36 |
-
"eval_runtime":
|
| 37 |
-
"eval_samples_per_second":
|
| 38 |
-
"eval_steps_per_second":
|
| 39 |
"step": 341
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0
|
| 43 |
-
"eval_bleu": 0.
|
| 44 |
-
"eval_exact_match": 0.
|
| 45 |
-
"eval_prefix_exact_match": 0.
|
| 46 |
"step": 341
|
| 47 |
},
|
| 48 |
{
|
| 49 |
-
"epoch": 1.
|
| 50 |
-
"grad_norm": 0.33546799421310425,
|
| 51 |
"learning_rate": 0.0001,
|
| 52 |
-
"loss": 1.
|
| 53 |
"step": 400
|
| 54 |
},
|
| 55 |
{
|
| 56 |
-
"epoch": 1.
|
| 57 |
-
"grad_norm": 0.3515666723251343,
|
| 58 |
"learning_rate": 0.0001,
|
| 59 |
-
"loss": 1.
|
| 60 |
"step": 500
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"epoch": 1.
|
| 64 |
-
"grad_norm": 0.35733896493911743,
|
| 65 |
"learning_rate": 0.0001,
|
| 66 |
-
"loss": 1.
|
| 67 |
"step": 600
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"epoch": 2.0,
|
| 71 |
-
"eval_accuracy": 0.
|
| 72 |
-
"eval_loss":
|
| 73 |
-
"eval_runtime":
|
| 74 |
-
"eval_samples_per_second":
|
| 75 |
-
"eval_steps_per_second":
|
| 76 |
"step": 683
|
| 77 |
},
|
| 78 |
{
|
| 79 |
"epoch": 2.0,
|
| 80 |
-
"eval_bleu": 0.
|
| 81 |
-
"eval_exact_match": 0.
|
| 82 |
-
"eval_prefix_exact_match": 0.
|
| 83 |
"step": 683
|
| 84 |
},
|
| 85 |
{
|
| 86 |
-
"epoch": 2.
|
| 87 |
-
"grad_norm": 0.3784458339214325,
|
| 88 |
"learning_rate": 0.0001,
|
| 89 |
-
"loss": 1.
|
| 90 |
"step": 700
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"epoch": 2.
|
| 94 |
-
"grad_norm": 0.46702849864959717,
|
| 95 |
"learning_rate": 0.0001,
|
| 96 |
-
"loss": 1.
|
| 97 |
"step": 800
|
| 98 |
},
|
| 99 |
{
|
| 100 |
-
"epoch": 2.
|
| 101 |
-
"grad_norm": 0.5156407356262207,
|
| 102 |
"learning_rate": 0.0001,
|
| 103 |
-
"loss": 1.
|
| 104 |
"step": 900
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch": 2.
|
| 108 |
-
"grad_norm": 0.4755038619041443,
|
| 109 |
"learning_rate": 0.0001,
|
| 110 |
-
"loss": 1.
|
| 111 |
"step": 1000
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"eval_accuracy": 0.
|
| 116 |
-
"eval_loss":
|
| 117 |
-
"eval_runtime":
|
| 118 |
-
"eval_samples_per_second":
|
| 119 |
-
"eval_steps_per_second":
|
| 120 |
"step": 1024
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"epoch":
|
| 124 |
-
"eval_bleu": 0.
|
| 125 |
-
"eval_exact_match": 0.
|
| 126 |
-
"eval_prefix_exact_match": 0.
|
| 127 |
"step": 1024
|
| 128 |
},
|
| 129 |
{
|
| 130 |
-
"epoch": 3.
|
| 131 |
-
"grad_norm": 0.5288150906562805,
|
| 132 |
"learning_rate": 0.0001,
|
| 133 |
-
"loss": 1.
|
| 134 |
"step": 1100
|
| 135 |
},
|
| 136 |
{
|
| 137 |
-
"epoch": 3.
|
| 138 |
-
"grad_norm": 0.5784010887145996,
|
| 139 |
"learning_rate": 0.0001,
|
| 140 |
-
"loss": 1.
|
| 141 |
"step": 1200
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 3.
|
| 145 |
-
"grad_norm": 0.6458429098129272,
|
| 146 |
"learning_rate": 0.0001,
|
| 147 |
-
"loss": 1.
|
| 148 |
"step": 1300
|
| 149 |
},
|
| 150 |
{
|
| 151 |
"epoch": 4.0,
|
| 152 |
-
"eval_accuracy": 0.
|
| 153 |
-
"eval_loss":
|
| 154 |
-
"eval_runtime": 4.
|
| 155 |
-
"eval_samples_per_second":
|
| 156 |
-
"eval_steps_per_second":
|
| 157 |
"step": 1366
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 4.0,
|
| 161 |
-
"eval_bleu": 0.
|
| 162 |
-
"eval_exact_match": 0.
|
| 163 |
-
"eval_prefix_exact_match": 0.
|
| 164 |
"step": 1366
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 4.
|
| 168 |
-
"grad_norm": 0.732841432094574,
|
| 169 |
"learning_rate": 0.0001,
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 1400
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 4.
|
| 175 |
-
"grad_norm": 0.7372816205024719,
|
| 176 |
"learning_rate": 0.0001,
|
| 177 |
-
"loss":
|
| 178 |
"step": 1500
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch": 4.
|
| 182 |
-
"grad_norm": 0.8235049247741699,
|
| 183 |
"learning_rate": 0.0001,
|
| 184 |
-
"loss": 1.
|
| 185 |
"step": 1600
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch": 4.
|
| 189 |
-
"grad_norm": 0.7765743732452393,
|
| 190 |
"learning_rate": 0.0001,
|
| 191 |
-
"loss":
|
| 192 |
"step": 1700
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch":
|
| 196 |
-
"eval_accuracy": 0.
|
| 197 |
-
"eval_loss":
|
| 198 |
-
"eval_runtime": 4.
|
| 199 |
-
"eval_samples_per_second":
|
| 200 |
-
"eval_steps_per_second": 12.
|
| 201 |
"step": 1707
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch":
|
| 205 |
-
"eval_bleu": 0.
|
| 206 |
-
"eval_exact_match": 0.
|
| 207 |
-
"eval_prefix_exact_match": 0.
|
| 208 |
"step": 1707
|
| 209 |
},
|
| 210 |
{
|
| 211 |
-
"epoch": 5.
|
| 212 |
-
"grad_norm": 0.9119400978088379,
|
| 213 |
"learning_rate": 0.0001,
|
| 214 |
-
"loss":
|
| 215 |
"step": 1800
|
| 216 |
},
|
| 217 |
{
|
| 218 |
-
"epoch": 5.
|
| 219 |
-
"grad_norm": 0.8966185450553894,
|
| 220 |
"learning_rate": 0.0001,
|
| 221 |
-
"loss":
|
| 222 |
"step": 1900
|
| 223 |
},
|
| 224 |
{
|
| 225 |
-
"epoch": 5.
|
| 226 |
-
"grad_norm": 0.9629669189453125,
|
| 227 |
"learning_rate": 0.0001,
|
| 228 |
-
"loss":
|
| 229 |
"step": 2000
|
| 230 |
},
|
| 231 |
{
|
| 232 |
"epoch": 6.0,
|
| 233 |
-
"eval_accuracy": 0.
|
| 234 |
-
"eval_loss":
|
| 235 |
-
"eval_runtime": 4.
|
| 236 |
-
"eval_samples_per_second":
|
| 237 |
-
"eval_steps_per_second":
|
| 238 |
"step": 2049
|
| 239 |
},
|
| 240 |
{
|
| 241 |
"epoch": 6.0,
|
| 242 |
-
"eval_bleu": 0.
|
| 243 |
-
"eval_exact_match": 0.
|
| 244 |
-
"eval_prefix_exact_match": 0.
|
| 245 |
"step": 2049
|
| 246 |
},
|
| 247 |
{
|
| 248 |
-
"epoch": 6.
|
| 249 |
-
"grad_norm": 1.0808535814285278,
|
| 250 |
"learning_rate": 0.0001,
|
| 251 |
-
"loss":
|
| 252 |
"step": 2100
|
| 253 |
},
|
| 254 |
{
|
| 255 |
-
"epoch": 6.
|
| 256 |
-
"grad_norm": 1.1639196872711182,
|
| 257 |
"learning_rate": 0.0001,
|
| 258 |
-
"loss":
|
| 259 |
"step": 2200
|
| 260 |
},
|
| 261 |
{
|
| 262 |
-
"epoch": 6.
|
| 263 |
-
"grad_norm": 1.0662291049957275,
|
| 264 |
"learning_rate": 0.0001,
|
| 265 |
-
"loss":
|
| 266 |
"step": 2300
|
| 267 |
},
|
| 268 |
{
|
| 269 |
-
"epoch":
|
| 270 |
-
"eval_accuracy": 0.
|
| 271 |
-
"eval_loss":
|
| 272 |
-
"eval_runtime":
|
| 273 |
-
"eval_samples_per_second":
|
| 274 |
-
"eval_steps_per_second":
|
| 275 |
"step": 2390
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch":
|
| 279 |
-
"eval_bleu": 0.17780932376070374,
|
| 280 |
-
"eval_exact_match": 0.004,
|
| 281 |
-
"eval_prefix_exact_match": 0.024,
|
| 282 |
"step": 2390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch": 7.
|
| 286 |
-
"grad_norm": 1.6462695598602295,
|
| 287 |
"learning_rate": 0.0001,
|
| 288 |
-
"loss":
|
| 289 |
"step": 2400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch": 7.
|
| 293 |
-
"grad_norm": 1.2542779445648193,
|
| 294 |
"learning_rate": 0.0001,
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 2500
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"epoch": 7.
|
| 300 |
-
"grad_norm": 1.334762454032898,
|
| 301 |
"learning_rate": 0.0001,
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 2600
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch": 7.
|
| 307 |
-
"grad_norm": 1.2401787042617798,
|
| 308 |
"learning_rate": 0.0001,
|
| 309 |
-
"loss": 0.
|
| 310 |
"step": 2700
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 8.0,
|
| 314 |
-
"eval_accuracy": 0.
|
| 315 |
-
"eval_loss":
|
| 316 |
-
"eval_runtime": 4.
|
| 317 |
-
"eval_samples_per_second":
|
| 318 |
-
"eval_steps_per_second":
|
| 319 |
"step": 2732
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 8.0,
|
| 323 |
-
"eval_bleu": 0.21418356891762502,
|
| 324 |
-
"eval_exact_match": 0.002,
|
| 325 |
-
"eval_prefix_exact_match": 0.026,
|
| 326 |
"step": 2732
|
| 327 |
},
|
| 328 |
{
|
| 329 |
-
"epoch": 8.
|
| 330 |
-
"grad_norm": 1.255996584892273,
|
| 331 |
"learning_rate": 0.0001,
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 2800
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch": 8.
|
| 337 |
-
"grad_norm": 1.2836525440216064,
|
| 338 |
"learning_rate": 0.0001,
|
| 339 |
-
"loss": 0.
|
| 340 |
"step": 2900
|
| 341 |
},
|
| 342 |
{
|
| 343 |
-
"epoch": 8.
|
| 344 |
-
"grad_norm": 1.375715732574463,
|
| 345 |
"learning_rate": 0.0001,
|
| 346 |
-
"loss": 0.
|
| 347 |
"step": 3000
|
| 348 |
},
|
| 349 |
{
|
| 350 |
-
"epoch":
|
| 351 |
-
"eval_accuracy": 0.
|
| 352 |
-
"eval_loss":
|
| 353 |
-
"eval_runtime":
|
| 354 |
-
"eval_samples_per_second":
|
| 355 |
-
"eval_steps_per_second":
|
| 356 |
"step": 3073
|
| 357 |
},
|
| 358 |
{
|
| 359 |
-
"epoch":
|
| 360 |
-
"eval_bleu": 0.20459866130314297,
|
| 361 |
-
"eval_exact_match": 0.0,
|
| 362 |
-
"eval_prefix_exact_match": 0.024,
|
| 363 |
"step": 3073
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch": 9.
|
| 367 |
-
"grad_norm": 1.481998324394226,
|
| 368 |
"learning_rate": 0.0001,
|
| 369 |
-
"loss": 0.
|
| 370 |
"step": 3100
|
| 371 |
},
|
| 372 |
{
|
| 373 |
-
"epoch": 9.
|
| 374 |
-
"grad_norm": 1.4086748361587524,
|
| 375 |
"learning_rate": 0.0001,
|
| 376 |
-
"loss": 0.
|
| 377 |
"step": 3200
|
| 378 |
},
|
| 379 |
{
|
| 380 |
-
"epoch": 9.
|
| 381 |
-
"grad_norm": 1.4381046295166016,
|
| 382 |
"learning_rate": 0.0001,
|
| 383 |
-
"loss": 0.
|
| 384 |
"step": 3300
|
| 385 |
},
|
| 386 |
{
|
| 387 |
-
"epoch": 9.
|
| 388 |
-
"grad_norm": 1.7364269495010376,
|
| 389 |
"learning_rate": 0.0001,
|
| 390 |
-
"loss": 0.
|
| 391 |
"step": 3400
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 10.0,
|
| 395 |
-
"eval_accuracy": 0.
|
| 396 |
-
"eval_loss":
|
| 397 |
-
"eval_runtime":
|
| 398 |
-
"eval_samples_per_second":
|
| 399 |
-
"eval_steps_per_second":
|
| 400 |
"step": 3415
|
| 401 |
},
|
| 402 |
{
|
| 403 |
"epoch": 10.0,
|
| 404 |
-
"eval_bleu": 0.19515580420261575,
|
| 405 |
-
"eval_exact_match": 0.002,
|
| 406 |
-
"eval_prefix_exact_match": 0.026,
|
| 407 |
"step": 3415
|
| 408 |
},
|
| 409 |
{
|
| 410 |
-
"epoch": 10.
|
| 411 |
-
"grad_norm": 1.599753737449646,
|
| 412 |
"learning_rate": 0.0001,
|
| 413 |
-
"loss": 0.
|
| 414 |
"step": 3500
|
| 415 |
},
|
| 416 |
{
|
| 417 |
-
"epoch": 10.
|
| 418 |
-
"grad_norm": 1.5595277547836304,
|
| 419 |
"learning_rate": 0.0001,
|
| 420 |
-
"loss": 0.
|
| 421 |
"step": 3600
|
| 422 |
},
|
| 423 |
{
|
| 424 |
-
"epoch": 10.
|
| 425 |
-
"grad_norm": 1.6201436519622803,
|
| 426 |
"learning_rate": 0.0001,
|
| 427 |
-
"loss": 0.
|
| 428 |
"step": 3700
|
| 429 |
},
|
| 430 |
{
|
| 431 |
-
"epoch":
|
| 432 |
-
"eval_accuracy": 0.
|
| 433 |
-
"eval_loss":
|
| 434 |
-
"eval_runtime":
|
| 435 |
-
"eval_samples_per_second":
|
| 436 |
-
"eval_steps_per_second":
|
| 437 |
"step": 3756
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch":
|
| 441 |
-
"eval_bleu": 0.2074466521823303,
|
| 442 |
-
"eval_exact_match": 0.0,
|
| 443 |
-
"eval_prefix_exact_match": 0.02,
|
| 444 |
"step": 3756
|
| 445 |
},
|
| 446 |
{
|
| 447 |
-
"epoch": 11.
|
| 448 |
-
"grad_norm": 1.6209241151809692,
|
| 449 |
"learning_rate": 0.0001,
|
| 450 |
-
"loss": 0.
|
| 451 |
"step": 3800
|
| 452 |
},
|
| 453 |
{
|
| 454 |
-
"epoch": 11.
|
| 455 |
-
"grad_norm": 1.5148394107818604,
|
| 456 |
"learning_rate": 0.0001,
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 3900
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch": 11.
|
| 462 |
-
"grad_norm": 1.6862399578094482,
|
| 463 |
"learning_rate": 0.0001,
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 4000
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 12.0,
|
| 469 |
-
"eval_accuracy": 0.
|
| 470 |
-
"eval_loss":
|
| 471 |
-
"eval_runtime":
|
| 472 |
-
"eval_samples_per_second":
|
| 473 |
-
"eval_steps_per_second":
|
| 474 |
"step": 4098
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 12.0,
|
| 478 |
-
"eval_bleu": 0.16189430124724263,
|
| 479 |
-
"eval_exact_match": 0.002,
|
| 480 |
-
"eval_prefix_exact_match": 0.028,
|
| 481 |
"step": 4098
|
| 482 |
},
|
| 483 |
{
|
| 484 |
-
"epoch": 12.
|
| 485 |
-
"grad_norm": 1.3296407461166382,
|
| 486 |
"learning_rate": 0.0001,
|
| 487 |
-
"loss": 0.
|
| 488 |
"step": 4100
|
| 489 |
},
|
| 490 |
{
|
| 491 |
-
"epoch": 12.
|
| 492 |
-
"grad_norm": 1.3439353704452515,
|
| 493 |
"learning_rate": 0.0001,
|
| 494 |
-
"loss": 0.
|
| 495 |
"step": 4200
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch": 12.
|
| 499 |
-
"grad_norm": 1.5329118967056274,
|
| 500 |
"learning_rate": 0.0001,
|
| 501 |
-
"loss": 0.
|
| 502 |
"step": 4300
|
| 503 |
},
|
| 504 |
{
|
| 505 |
-
"epoch": 12.
|
| 506 |
-
"grad_norm": 1.649775743484497,
|
| 507 |
"learning_rate": 0.0001,
|
| 508 |
-
"loss": 0.
|
| 509 |
"step": 4400
|
| 510 |
},
|
| 511 |
{
|
| 512 |
-
"epoch":
|
| 513 |
-
"eval_accuracy": 0.
|
| 514 |
-
"eval_loss":
|
| 515 |
-
"eval_runtime":
|
| 516 |
-
"eval_samples_per_second":
|
| 517 |
-
"eval_steps_per_second":
|
| 518 |
"step": 4439
|
| 519 |
},
|
| 520 |
{
|
| 521 |
-
"epoch":
|
| 522 |
-
"eval_bleu": 0.17170793542335236,
|
| 523 |
-
"eval_exact_match": 0.002,
|
| 524 |
-
"eval_prefix_exact_match": 0.028,
|
| 525 |
"step": 4439
|
| 526 |
},
|
| 527 |
{
|
| 528 |
-
"epoch": 13.
|
| 529 |
-
"grad_norm": 1.5262281894683838,
|
| 530 |
"learning_rate": 0.0001,
|
| 531 |
-
"loss": 0.
|
| 532 |
"step": 4500
|
| 533 |
},
|
| 534 |
{
|
| 535 |
-
"epoch": 13.
|
| 536 |
-
"grad_norm": 1.5962131023406982,
|
| 537 |
"learning_rate": 0.0001,
|
| 538 |
-
"loss": 0.
|
| 539 |
"step": 4600
|
| 540 |
},
|
| 541 |
{
|
| 542 |
-
"epoch": 13.
|
| 543 |
-
"grad_norm": 1.4258382320404053,
|
| 544 |
"learning_rate": 0.0001,
|
| 545 |
-
"loss": 0.
|
| 546 |
"step": 4700
|
| 547 |
},
|
| 548 |
{
|
| 549 |
"epoch": 14.0,
|
| 550 |
-
"eval_accuracy": 0.
|
| 551 |
-
"eval_loss":
|
| 552 |
-
"eval_runtime": 4.
|
| 553 |
-
"eval_samples_per_second":
|
| 554 |
-
"eval_steps_per_second":
|
| 555 |
"step": 4781
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 14.0,
|
| 559 |
-
"eval_bleu": 0.
|
| 560 |
-
"eval_exact_match": 0.
|
| 561 |
-
"eval_prefix_exact_match": 0.
|
| 562 |
"step": 4781
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"epoch": 14.
|
| 566 |
-
"grad_norm": 1.388717532157898,
|
| 567 |
"learning_rate": 0.0001,
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 4800
|
| 570 |
},
|
| 571 |
{
|
| 572 |
-
"epoch": 14.
|
| 573 |
-
"grad_norm": 1.2370668649673462,
|
| 574 |
"learning_rate": 0.0001,
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 4900
|
| 577 |
},
|
| 578 |
{
|
| 579 |
-
"epoch": 14.
|
| 580 |
-
"grad_norm": 1.3778610229492188,
|
| 581 |
"learning_rate": 0.0001,
|
| 582 |
-
"loss": 0.
|
| 583 |
"step": 5000
|
| 584 |
},
|
| 585 |
{
|
| 586 |
-
"epoch": 14.
|
| 587 |
-
"grad_norm": 1.6371277570724487,
|
| 588 |
"learning_rate": 0.0001,
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 5100
|
| 591 |
},
|
| 592 |
{
|
| 593 |
-
"epoch":
|
| 594 |
-
"eval_accuracy": 0.
|
| 595 |
-
"eval_loss":
|
| 596 |
-
"eval_runtime": 5.
|
| 597 |
-
"eval_samples_per_second":
|
| 598 |
-
"eval_steps_per_second": 12.
|
| 599 |
"step": 5122
|
| 600 |
},
|
| 601 |
{
|
| 602 |
-
"epoch":
|
| 603 |
-
"eval_bleu": 0.
|
| 604 |
"eval_exact_match": 0.0,
|
| 605 |
-
"eval_prefix_exact_match": 0.
|
| 606 |
"step": 5122
|
| 607 |
},
|
| 608 |
{
|
| 609 |
-
"epoch": 15.
|
| 610 |
-
"grad_norm": 1.2761411666870117,
|
| 611 |
"learning_rate": 0.0001,
|
| 612 |
-
"loss": 0.
|
| 613 |
"step": 5200
|
| 614 |
},
|
| 615 |
{
|
| 616 |
-
"epoch": 15.
|
| 617 |
-
"grad_norm": 1.6878156661987305,
|
| 618 |
"learning_rate": 0.0001,
|
| 619 |
-
"loss": 0.
|
| 620 |
"step": 5300
|
| 621 |
},
|
| 622 |
{
|
| 623 |
-
"epoch": 15.
|
| 624 |
-
"grad_norm": 1.4964863061904907,
|
| 625 |
"learning_rate": 0.0001,
|
| 626 |
-
"loss": 0.
|
| 627 |
"step": 5400
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 16.0,
|
| 631 |
-
"eval_accuracy": 0.
|
| 632 |
-
"eval_loss":
|
| 633 |
-
"eval_runtime": 4.
|
| 634 |
-
"eval_samples_per_second":
|
| 635 |
-
"eval_steps_per_second":
|
| 636 |
"step": 5464
|
| 637 |
},
|
| 638 |
{
|
| 639 |
"epoch": 16.0,
|
| 640 |
-
"eval_bleu": 0.
|
| 641 |
-
"eval_exact_match": 0.
|
| 642 |
-
"eval_prefix_exact_match": 0.
|
| 643 |
"step": 5464
|
| 644 |
},
|
| 645 |
{
|
| 646 |
-
"epoch": 16.
|
| 647 |
-
"grad_norm": 0.9965784549713135,
|
| 648 |
"learning_rate": 0.0001,
|
| 649 |
-
"loss": 0.
|
| 650 |
"step": 5500
|
| 651 |
},
|
| 652 |
{
|
| 653 |
-
"epoch": 16.
|
| 654 |
-
"grad_norm": 1.2694511413574219,
|
| 655 |
"learning_rate": 0.0001,
|
| 656 |
-
"loss": 0.
|
| 657 |
"step": 5600
|
| 658 |
},
|
| 659 |
{
|
| 660 |
-
"epoch": 16.
|
| 661 |
-
"grad_norm": 1.436286211013794,
|
| 662 |
"learning_rate": 0.0001,
|
| 663 |
-
"loss": 0.
|
| 664 |
"step": 5700
|
| 665 |
},
|
| 666 |
{
|
| 667 |
-
"epoch": 16.
|
| 668 |
-
"grad_norm": 1.526081919670105,
|
| 669 |
"learning_rate": 0.0001,
|
| 670 |
-
"loss": 0.
|
| 671 |
"step": 5800
|
| 672 |
},
|
| 673 |
{
|
| 674 |
-
"epoch":
|
| 675 |
-
"eval_accuracy": 0.
|
| 676 |
-
"eval_loss":
|
| 677 |
-
"eval_runtime":
|
| 678 |
-
"eval_samples_per_second":
|
| 679 |
-
"eval_steps_per_second":
|
| 680 |
"step": 5805
|
| 681 |
},
|
| 682 |
{
|
| 683 |
-
"epoch":
|
| 684 |
-
"eval_bleu": 0.
|
| 685 |
"eval_exact_match": 0.0,
|
| 686 |
-
"eval_prefix_exact_match": 0.
|
| 687 |
"step": 5805
|
| 688 |
},
|
| 689 |
{
|
| 690 |
-
"epoch": 17.
|
| 691 |
-
"grad_norm": 1.1292760372161865,
|
| 692 |
"learning_rate": 0.0001,
|
| 693 |
-
"loss": 0.
|
| 694 |
"step": 5900
|
| 695 |
},
|
| 696 |
{
|
| 697 |
-
"epoch": 17.
|
| 698 |
-
"grad_norm": 1.440766453742981,
|
| 699 |
"learning_rate": 0.0001,
|
| 700 |
-
"loss": 0.
|
| 701 |
"step": 6000
|
| 702 |
},
|
| 703 |
{
|
| 704 |
-
"epoch": 17.
|
| 705 |
-
"grad_norm": 1.2640533447265625,
|
| 706 |
"learning_rate": 0.0001,
|
| 707 |
-
"loss": 0.
|
| 708 |
"step": 6100
|
| 709 |
},
|
| 710 |
{
|
| 711 |
"epoch": 18.0,
|
| 712 |
-
"eval_accuracy": 0.
|
| 713 |
-
"eval_loss":
|
| 714 |
-
"eval_runtime":
|
| 715 |
-
"eval_samples_per_second":
|
| 716 |
-
"eval_steps_per_second":
|
| 717 |
"step": 6147
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 18.0,
|
| 721 |
-
"eval_bleu": 0.
|
| 722 |
"eval_exact_match": 0.0,
|
| 723 |
-
"eval_prefix_exact_match": 0.
|
| 724 |
"step": 6147
|
| 725 |
},
|
| 726 |
{
|
| 727 |
-
"epoch": 18.
|
| 728 |
-
"grad_norm": 1.430617094039917,
|
| 729 |
"learning_rate": 0.0001,
|
| 730 |
-
"loss": 0.
|
| 731 |
"step": 6200
|
| 732 |
},
|
| 733 |
{
|
| 734 |
-
"epoch": 18.
|
| 735 |
-
"grad_norm": 1.1885381937026978,
|
| 736 |
"learning_rate": 0.0001,
|
| 737 |
-
"loss": 0.
|
| 738 |
"step": 6300
|
| 739 |
},
|
| 740 |
{
|
| 741 |
-
"epoch": 18.
|
| 742 |
-
"grad_norm": 1.64545738697052,
|
| 743 |
"learning_rate": 0.0001,
|
| 744 |
-
"loss": 0.
|
| 745 |
"step": 6400
|
| 746 |
},
|
| 747 |
{
|
| 748 |
-
"epoch":
|
| 749 |
-
"eval_accuracy": 0.
|
| 750 |
-
"eval_loss":
|
| 751 |
-
"eval_runtime":
|
| 752 |
-
"eval_samples_per_second":
|
| 753 |
-
"eval_steps_per_second":
|
| 754 |
"step": 6488
|
| 755 |
},
|
| 756 |
{
|
| 757 |
-
"epoch":
|
| 758 |
-
"eval_bleu": 0.
|
| 759 |
"eval_exact_match": 0.0,
|
| 760 |
-
"eval_prefix_exact_match": 0.
|
| 761 |
"step": 6488
|
| 762 |
},
|
| 763 |
{
|
| 764 |
-
"epoch": 19.
|
| 765 |
-
"grad_norm": 1.2168452739715576,
|
| 766 |
"learning_rate": 0.0001,
|
| 767 |
-
"loss": 0.
|
| 768 |
"step": 6500
|
| 769 |
},
|
| 770 |
{
|
| 771 |
-
"epoch": 19.
|
| 772 |
-
"grad_norm": 1.2320917844772339,
|
| 773 |
"learning_rate": 0.0001,
|
| 774 |
-
"loss": 0.
|
| 775 |
"step": 6600
|
| 776 |
},
|
| 777 |
{
|
| 778 |
-
"epoch": 19.
|
| 779 |
-
"grad_norm": 1.1541388034820557,
|
| 780 |
"learning_rate": 0.0001,
|
| 781 |
-
"loss": 0.
|
| 782 |
"step": 6700
|
| 783 |
},
|
| 784 |
{
|
| 785 |
-
"epoch": 19.
|
| 786 |
-
"grad_norm": 1.2821346521377563,
|
| 787 |
"learning_rate": 0.0001,
|
| 788 |
-
"loss": 0.
|
| 789 |
"step": 6800
|
| 790 |
},
|
| 791 |
{
|
| 792 |
"epoch": 20.0,
|
| 793 |
-
"eval_accuracy": 0.
|
| 794 |
-
"eval_loss":
|
| 795 |
-
"eval_runtime":
|
| 796 |
-
"eval_samples_per_second":
|
| 797 |
-
"eval_steps_per_second":
|
| 798 |
"step": 6830
|
| 799 |
},
|
| 800 |
{
|
| 801 |
"epoch": 20.0,
|
| 802 |
-
"eval_bleu": 0.
|
| 803 |
"eval_exact_match": 0.0,
|
| 804 |
-
"eval_prefix_exact_match": 0.
|
| 805 |
"step": 6830
|
| 806 |
},
|
| 807 |
{
|
| 808 |
-
"epoch": 20.
|
| 809 |
-
"grad_norm": 1.1950445175170898,
|
| 810 |
"learning_rate": 0.0001,
|
| 811 |
-
"loss": 0.
|
| 812 |
"step": 6900
|
| 813 |
},
|
| 814 |
{
|
| 815 |
-
"epoch": 20.
|
| 816 |
-
"grad_norm": 1.253625750541687,
|
| 817 |
"learning_rate": 0.0001,
|
| 818 |
-
"loss": 0.
|
| 819 |
"step": 7000
|
| 820 |
},
|
| 821 |
{
|
| 822 |
-
"epoch": 20.
|
| 823 |
-
"grad_norm": 1.193354606628418,
|
| 824 |
"learning_rate": 0.0001,
|
| 825 |
-
"loss": 0.
|
| 826 |
"step": 7100
|
| 827 |
},
|
| 828 |
{
|
| 829 |
-
"epoch":
|
| 830 |
-
"eval_accuracy": 0.
|
| 831 |
-
"eval_loss":
|
| 832 |
-
"eval_runtime":
|
| 833 |
-
"eval_samples_per_second":
|
| 834 |
-
"eval_steps_per_second":
|
| 835 |
"step": 7171
|
| 836 |
},
|
| 837 |
{
|
| 838 |
-
"epoch":
|
| 839 |
-
"eval_bleu": 0.16340256740662695,
|
| 840 |
-
"eval_exact_match": 0.0,
|
| 841 |
-
"eval_prefix_exact_match": 0.022,
|
| 842 |
"step": 7171
|
| 843 |
},
|
| 844 |
{
|
| 845 |
-
"epoch": 21.
|
| 846 |
-
"grad_norm": 1.0565904378890991,
|
| 847 |
"learning_rate": 0.0001,
|
| 848 |
-
"loss": 0.
|
| 849 |
"step": 7200
|
| 850 |
},
|
| 851 |
{
|
| 852 |
-
"epoch": 21.
|
| 853 |
-
"grad_norm": 1.1256818771362305,
|
| 854 |
"learning_rate": 0.0001,
|
| 855 |
-
"loss": 0.
|
| 856 |
"step": 7300
|
| 857 |
},
|
| 858 |
{
|
| 859 |
-
"epoch": 21.
|
| 860 |
-
"grad_norm": 1.4385403394699097,
|
| 861 |
"learning_rate": 0.0001,
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 7400
|
| 864 |
},
|
| 865 |
{
|
| 866 |
-
"epoch": 21.
|
| 867 |
-
"grad_norm": 1.1152088642120361,
|
| 868 |
"learning_rate": 0.0001,
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 7500
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 22.0,
|
| 874 |
-
"eval_accuracy": 0.
|
| 875 |
-
"eval_loss":
|
| 876 |
-
"eval_runtime":
|
| 877 |
-
"eval_samples_per_second":
|
| 878 |
-
"eval_steps_per_second":
|
| 879 |
"step": 7513
|
| 880 |
},
|
| 881 |
{
|
| 882 |
"epoch": 22.0,
|
| 883 |
-
"eval_bleu": 0.19792591482411717,
|
| 884 |
-
"eval_exact_match": 0.0,
|
| 885 |
-
"eval_prefix_exact_match": 0.022,
|
| 886 |
"step": 7513
|
| 887 |
},
|
| 888 |
{
|
| 889 |
-
"epoch": 22.
|
| 890 |
-
"grad_norm": 1.1440153121948242,
|
| 891 |
"learning_rate": 0.0001,
|
| 892 |
-
"loss": 0.
|
| 893 |
"step": 7600
|
| 894 |
},
|
| 895 |
{
|
| 896 |
-
"epoch": 22.
|
| 897 |
-
"grad_norm": 1.1086493730545044,
|
| 898 |
"learning_rate": 0.0001,
|
| 899 |
-
"loss": 0.
|
| 900 |
"step": 7700
|
| 901 |
},
|
| 902 |
{
|
| 903 |
-
"epoch": 22.
|
| 904 |
-
"grad_norm": 1.2570273876190186,
|
| 905 |
"learning_rate": 0.0001,
|
| 906 |
-
"loss": 0.
|
| 907 |
"step": 7800
|
| 908 |
},
|
| 909 |
{
|
| 910 |
-
"epoch":
|
| 911 |
-
"eval_accuracy": 0.
|
| 912 |
-
"eval_loss":
|
| 913 |
-
"eval_runtime":
|
| 914 |
-
"eval_samples_per_second":
|
| 915 |
-
"eval_steps_per_second":
|
| 916 |
"step": 7854
|
| 917 |
},
|
| 918 |
{
|
| 919 |
-
"epoch":
|
| 920 |
-
"eval_bleu": 0.
|
| 921 |
-
"eval_exact_match": 0.
|
| 922 |
-
"eval_prefix_exact_match": 0.
|
| 923 |
"step": 7854
|
| 924 |
},
|
| 925 |
{
|
| 926 |
-
"epoch": 23.
|
| 927 |
-
"grad_norm": 1.070988655090332,
|
| 928 |
"learning_rate": 0.0001,
|
| 929 |
-
"loss": 0.
|
| 930 |
"step": 7900
|
| 931 |
},
|
| 932 |
{
|
| 933 |
-
"epoch": 23.
|
| 934 |
-
"grad_norm": 1.021813154220581,
|
| 935 |
"learning_rate": 0.0001,
|
| 936 |
-
"loss": 0.
|
| 937 |
"step": 8000
|
| 938 |
},
|
| 939 |
{
|
| 940 |
-
"epoch": 23.
|
| 941 |
-
"grad_norm": 1.2333232164382935,
|
| 942 |
"learning_rate": 0.0001,
|
| 943 |
-
"loss": 0.
|
| 944 |
"step": 8100
|
| 945 |
},
|
| 946 |
{
|
| 947 |
"epoch": 24.0,
|
| 948 |
-
"eval_accuracy": 0.
|
| 949 |
-
"eval_loss":
|
| 950 |
-
"eval_runtime": 4.
|
| 951 |
-
"eval_samples_per_second":
|
| 952 |
-
"eval_steps_per_second":
|
| 953 |
"step": 8196
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 24.0,
|
| 957 |
-
"eval_bleu": 0.1531918581052099,
|
| 958 |
-
"eval_exact_match": 0.0,
|
| 959 |
-
"eval_prefix_exact_match": 0.028,
|
| 960 |
"step": 8196
|
| 961 |
},
|
| 962 |
{
|
| 963 |
-
"epoch": 24.
|
| 964 |
-
"grad_norm": 0.9755038022994995,
|
| 965 |
"learning_rate": 0.0001,
|
| 966 |
-
"loss": 0.
|
| 967 |
"step": 8200
|
| 968 |
},
|
| 969 |
{
|
| 970 |
-
"epoch": 24.
|
| 971 |
-
"grad_norm": 1.2587764263153076,
|
| 972 |
"learning_rate": 0.0001,
|
| 973 |
-
"loss": 0.
|
| 974 |
"step": 8300
|
| 975 |
},
|
| 976 |
{
|
| 977 |
-
"epoch": 24.
|
| 978 |
-
"grad_norm": 0.9521226286888123,
|
| 979 |
"learning_rate": 0.0001,
|
| 980 |
-
"loss": 0.
|
| 981 |
"step": 8400
|
| 982 |
},
|
| 983 |
{
|
| 984 |
-
"epoch": 24.
|
| 985 |
-
"grad_norm": 1.2158619165420532,
|
| 986 |
"learning_rate": 0.0001,
|
| 987 |
-
"loss": 0.
|
| 988 |
"step": 8500
|
| 989 |
},
|
| 990 |
{
|
| 991 |
-
"epoch":
|
| 992 |
-
"eval_accuracy": 0.
|
| 993 |
-
"eval_loss":
|
| 994 |
-
"eval_runtime":
|
| 995 |
-
"eval_samples_per_second":
|
| 996 |
-
"eval_steps_per_second":
|
| 997 |
"step": 8537
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
-
"epoch":
|
| 1001 |
-
"eval_bleu": 0.17454308300545518,
|
| 1002 |
-
"eval_exact_match": 0.0,
|
| 1003 |
-
"eval_prefix_exact_match": 0.022,
|
| 1004 |
"step": 8537
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
-
"epoch": 25.
|
| 1008 |
-
"grad_norm": 1.0446624755859375,
|
| 1009 |
"learning_rate": 0.0001,
|
| 1010 |
-
"loss": 0.
|
| 1011 |
"step": 8600
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
-
"epoch": 25.
|
| 1015 |
-
"grad_norm": 1.07607901096344,
|
| 1016 |
"learning_rate": 0.0001,
|
| 1017 |
-
"loss": 0.
|
| 1018 |
"step": 8700
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
-
"epoch": 25.
|
| 1022 |
-
"grad_norm": 1.0376451015472412,
|
| 1023 |
"learning_rate": 0.0001,
|
| 1024 |
-
"loss": 0.
|
| 1025 |
"step": 8800
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 26.0,
|
| 1029 |
-
"eval_accuracy": 0.
|
| 1030 |
-
"eval_loss":
|
| 1031 |
-
"eval_runtime":
|
| 1032 |
-
"eval_samples_per_second":
|
| 1033 |
-
"eval_steps_per_second":
|
| 1034 |
"step": 8879
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
"epoch": 26.0,
|
| 1038 |
-
"eval_bleu": 0.183512215739427,
|
| 1039 |
-
"eval_exact_match": 0.0,
|
| 1040 |
-
"eval_prefix_exact_match": 0.026,
|
| 1041 |
"step": 8879
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
-
"epoch": 26.
|
| 1045 |
-
"grad_norm": 0.8689504861831665,
|
| 1046 |
"learning_rate": 0.0001,
|
| 1047 |
-
"loss": 0.
|
| 1048 |
"step": 8900
|
| 1049 |
},
|
| 1050 |
{
|
| 1051 |
-
"epoch": 26.
|
| 1052 |
-
"grad_norm": 0.8533344864845276,
|
| 1053 |
"learning_rate": 0.0001,
|
| 1054 |
-
"loss": 0.
|
| 1055 |
"step": 9000
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
-
"epoch": 26.
|
| 1059 |
-
"grad_norm": 1.0425468683242798,
|
| 1060 |
"learning_rate": 0.0001,
|
| 1061 |
-
"loss": 0.
|
| 1062 |
"step": 9100
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
-
"epoch": 26.
|
| 1066 |
-
"grad_norm": 1.2276948690414429,
|
| 1067 |
"learning_rate": 0.0001,
|
| 1068 |
-
"loss": 0.
|
| 1069 |
"step": 9200
|
| 1070 |
},
|
| 1071 |
{
|
| 1072 |
-
"epoch":
|
| 1073 |
-
"eval_accuracy": 0.
|
| 1074 |
-
"eval_loss":
|
| 1075 |
-
"eval_runtime":
|
| 1076 |
-
"eval_samples_per_second":
|
| 1077 |
-
"eval_steps_per_second":
|
| 1078 |
"step": 9220
|
| 1079 |
},
|
| 1080 |
{
|
| 1081 |
-
"epoch":
|
| 1082 |
-
"eval_bleu": 0.19671240169357163,
|
| 1083 |
-
"eval_exact_match": 0.002,
|
| 1084 |
-
"eval_prefix_exact_match": 0.026,
|
| 1085 |
"step": 9220
|
| 1086 |
},
|
| 1087 |
{
|
| 1088 |
-
"epoch": 27.
|
| 1089 |
-
"grad_norm": 1.0451931953430176,
|
| 1090 |
"learning_rate": 0.0001,
|
| 1091 |
-
"loss": 0.
|
| 1092 |
"step": 9300
|
| 1093 |
},
|
| 1094 |
{
|
| 1095 |
-
"epoch": 27.
|
| 1096 |
-
"grad_norm": 1.0023565292358398,
|
| 1097 |
"learning_rate": 0.0001,
|
| 1098 |
-
"loss": 0.
|
| 1099 |
"step": 9400
|
| 1100 |
},
|
| 1101 |
{
|
| 1102 |
-
"epoch": 27.
|
| 1103 |
-
"grad_norm": 0.9287552833557129,
|
| 1104 |
"learning_rate": 0.0001,
|
| 1105 |
-
"loss": 0.
|
| 1106 |
"step": 9500
|
| 1107 |
},
|
| 1108 |
{
|
| 1109 |
"epoch": 28.0,
|
| 1110 |
-
"eval_accuracy": 0.
|
| 1111 |
-
"eval_loss":
|
| 1112 |
-
"eval_runtime": 4.
|
| 1113 |
-
"eval_samples_per_second":
|
| 1114 |
-
"eval_steps_per_second":
|
| 1115 |
"step": 9562
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 28.0,
|
| 1119 |
-
"eval_bleu": 0.17149559728372338,
|
| 1120 |
-
"eval_exact_match": 0.0,
|
| 1121 |
-
"eval_prefix_exact_match": 0.028,
|
| 1122 |
"step": 9562
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
-
"epoch": 28.
|
| 1126 |
-
"grad_norm": 0.7279195785522461,
|
| 1127 |
"learning_rate": 0.0001,
|
| 1128 |
-
"loss": 0.
|
| 1129 |
"step": 9600
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
-
"epoch": 28.
|
| 1133 |
-
"grad_norm": 0.9348853230476379,
|
| 1134 |
"learning_rate": 0.0001,
|
| 1135 |
-
"loss": 0.
|
| 1136 |
"step": 9700
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
-
"epoch": 28.
|
| 1140 |
-
"grad_norm": 0.9402393102645874,
|
| 1141 |
"learning_rate": 0.0001,
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 9800
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
-
"epoch": 28.
|
| 1147 |
-
"grad_norm": 0.9316591620445251,
|
| 1148 |
"learning_rate": 0.0001,
|
| 1149 |
-
"loss": 0.
|
| 1150 |
"step": 9900
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
-
"epoch":
|
| 1154 |
-
"eval_accuracy": 0.
|
| 1155 |
-
"eval_loss":
|
| 1156 |
-
"eval_runtime":
|
| 1157 |
-
"eval_samples_per_second":
|
| 1158 |
-
"eval_steps_per_second":
|
| 1159 |
"step": 9903
|
| 1160 |
},
|
| 1161 |
{
|
| 1162 |
-
"epoch":
|
| 1163 |
-
"eval_bleu": 0.16884007365872924,
|
| 1164 |
-
"eval_exact_match": 0.0,
|
| 1165 |
-
"eval_prefix_exact_match": 0.02,
|
| 1166 |
"step": 9903
|
| 1167 |
},
|
| 1168 |
{
|
| 1169 |
-
"epoch": 29.
|
| 1170 |
-
"grad_norm": 0.9301176071166992,
|
| 1171 |
"learning_rate": 0.0001,
|
| 1172 |
-
"loss": 0.
|
| 1173 |
"step": 10000
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
-
"epoch": 29.
|
| 1177 |
-
"grad_norm": 0.8069124221801758,
|
| 1178 |
"learning_rate": 0.0001,
|
| 1179 |
-
"loss": 0.
|
| 1180 |
"step": 10100
|
| 1181 |
},
|
| 1182 |
{
|
| 1183 |
-
"epoch": 29.
|
| 1184 |
-
"grad_norm": 1.0093735456466675,
|
| 1185 |
"learning_rate": 0.0001,
|
| 1186 |
-
"loss": 0.
|
| 1187 |
"step": 10200
|
| 1188 |
},
|
| 1189 |
{
|
| 1190 |
"epoch": 30.0,
|
| 1191 |
-
"eval_accuracy": 0.
|
| 1192 |
-
"eval_loss":
|
| 1193 |
-
"eval_runtime": 4.
|
| 1194 |
-
"eval_samples_per_second":
|
| 1195 |
-
"eval_steps_per_second": 13.
|
| 1196 |
"step": 10245
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 30.0,
|
| 1200 |
-
"eval_bleu": 0.19612712097933296,
|
| 1201 |
-
"eval_exact_match": 0.0,
|
| 1202 |
-
"eval_prefix_exact_match": 0.028,
|
| 1203 |
"step": 10245
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
-
"epoch": 30.
|
| 1207 |
-
"grad_norm": 0.965752899646759,
|
| 1208 |
"learning_rate": 0.0001,
|
| 1209 |
-
"loss": 0.
|
| 1210 |
"step": 10300
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
-
"epoch": 30.
|
| 1214 |
-
"grad_norm": 0.9576678276062012,
|
| 1215 |
"learning_rate": 0.0001,
|
| 1216 |
-
"loss": 0.
|
| 1217 |
"step": 10400
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
-
"epoch": 30.
|
| 1221 |
-
"grad_norm": 0.8335288763046265,
|
| 1222 |
"learning_rate": 0.0001,
|
| 1223 |
-
"loss": 0.
|
| 1224 |
"step": 10500
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
-
"epoch":
|
| 1228 |
-
"eval_accuracy": 0.
|
| 1229 |
-
"eval_loss":
|
| 1230 |
-
"eval_runtime":
|
| 1231 |
-
"eval_samples_per_second":
|
| 1232 |
-
"eval_steps_per_second":
|
| 1233 |
"step": 10586
|
| 1234 |
},
|
| 1235 |
{
|
| 1236 |
-
"epoch":
|
| 1237 |
-
"eval_bleu": 0.17063063536886222,
|
| 1238 |
-
"eval_exact_match": 0.0,
|
| 1239 |
-
"eval_prefix_exact_match": 0.026,
|
| 1240 |
"step": 10586
|
| 1241 |
}
|
| 1242 |
],
|
| 1243 |
"logging_steps": 100,
|
| 1244 |
"max_steps": 17050,
|
| 1245 |
-
"num_input_tokens_seen": 0,
|
| 1246 |
"num_train_epochs": 50,
|
| 1247 |
"save_steps": 500,
|
| 1248 |
-
"
|
| 1249 |
-
"TrainerControl": {
|
| 1250 |
-
"args": {
|
| 1251 |
-
"should_epoch_stop": false,
|
| 1252 |
-
"should_evaluate": false,
|
| 1253 |
-
"should_log": false,
|
| 1254 |
-
"should_save": true,
|
| 1255 |
-
"should_training_stop": false
|
| 1256 |
-
},
|
| 1257 |
-
"attributes": {}
|
| 1258 |
-
}
|
| 1259 |
-
},
|
| 1260 |
-
"total_flos": 9.090794837749268e+17,
|
| 1261 |
-
"train_batch_size": 1,
|
| 1262 |
"trial_name": null,
|
| 1263 |
"trial_params": null
|
| 1264 |
}
|
|
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.29,
|
|
|
|
| 13 |
"learning_rate": 0.0001,
|
| 14 |
+
"loss": 1.4842,
|
| 15 |
"step": 100
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.59,
|
|
|
|
| 19 |
"learning_rate": 0.0001,
|
| 20 |
+
"loss": 1.4016,
|
| 21 |
"step": 200
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.88,
|
|
|
|
| 25 |
"learning_rate": 0.0001,
|
| 26 |
+
"loss": 1.3892,
|
| 27 |
"step": 300
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 1.0,
|
| 31 |
+
"eval_accuracy": 0.4543589743589744,
|
| 32 |
+
"eval_loss": 3.4055869579315186,
|
| 33 |
+
"eval_runtime": 4.1688,
|
| 34 |
+
"eval_samples_per_second": 119.938,
|
| 35 |
+
"eval_steps_per_second": 15.112,
|
| 36 |
"step": 341
|
| 37 |
},
|
| 38 |
{
|
| 39 |
+
"epoch": 1.0,
|
| 40 |
+
"eval_bleu": 0.09711501224106496,
|
| 41 |
+
"eval_exact_match": 0.0,
|
| 42 |
+
"eval_prefix_exact_match": 0.054,
|
| 43 |
"step": 341
|
| 44 |
},
|
| 45 |
{
|
| 46 |
+
"epoch": 1.17,
|
|
|
|
| 47 |
"learning_rate": 0.0001,
|
| 48 |
+
"loss": 1.3608,
|
| 49 |
"step": 400
|
| 50 |
},
|
| 51 |
{
|
| 52 |
+
"epoch": 1.46,
|
|
|
|
| 53 |
"learning_rate": 0.0001,
|
| 54 |
+
"loss": 1.3456,
|
| 55 |
"step": 500
|
| 56 |
},
|
| 57 |
{
|
| 58 |
+
"epoch": 1.76,
|
|
|
|
| 59 |
"learning_rate": 0.0001,
|
| 60 |
+
"loss": 1.3499,
|
| 61 |
"step": 600
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 2.0,
|
| 65 |
+
"eval_accuracy": 0.4576923076923077,
|
| 66 |
+
"eval_loss": 3.453113317489624,
|
| 67 |
+
"eval_runtime": 4.4031,
|
| 68 |
+
"eval_samples_per_second": 113.556,
|
| 69 |
+
"eval_steps_per_second": 14.308,
|
| 70 |
"step": 683
|
| 71 |
},
|
| 72 |
{
|
| 73 |
"epoch": 2.0,
|
| 74 |
+
"eval_bleu": 0.1225593574040313,
|
| 75 |
+
"eval_exact_match": 0.0,
|
| 76 |
+
"eval_prefix_exact_match": 0.056,
|
| 77 |
"step": 683
|
| 78 |
},
|
| 79 |
{
|
| 80 |
+
"epoch": 2.05,
|
|
|
|
| 81 |
"learning_rate": 0.0001,
|
| 82 |
+
"loss": 1.3238,
|
| 83 |
"step": 700
|
| 84 |
},
|
| 85 |
{
|
| 86 |
+
"epoch": 2.34,
|
|
|
|
| 87 |
"learning_rate": 0.0001,
|
| 88 |
+
"loss": 1.2486,
|
| 89 |
"step": 800
|
| 90 |
},
|
| 91 |
{
|
| 92 |
+
"epoch": 2.64,
|
|
|
|
| 93 |
"learning_rate": 0.0001,
|
| 94 |
+
"loss": 1.2452,
|
| 95 |
"step": 900
|
| 96 |
},
|
| 97 |
{
|
| 98 |
+
"epoch": 2.93,
|
|
|
|
| 99 |
"learning_rate": 0.0001,
|
| 100 |
+
"loss": 1.2427,
|
| 101 |
"step": 1000
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 3.0,
|
| 105 |
+
"eval_accuracy": 0.4584102564102564,
|
| 106 |
+
"eval_loss": 3.671137809753418,
|
| 107 |
+
"eval_runtime": 4.4094,
|
| 108 |
+
"eval_samples_per_second": 113.393,
|
| 109 |
+
"eval_steps_per_second": 14.288,
|
| 110 |
"step": 1024
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"epoch": 3.0,
|
| 114 |
+
"eval_bleu": 0.13165889774889974,
|
| 115 |
+
"eval_exact_match": 0.0,
|
| 116 |
+
"eval_prefix_exact_match": 0.056,
|
| 117 |
"step": 1024
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 3.22,
|
|
|
|
| 121 |
"learning_rate": 0.0001,
|
| 122 |
+
"loss": 1.1507,
|
| 123 |
"step": 1100
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 3.51,
|
|
|
|
| 127 |
"learning_rate": 0.0001,
|
| 128 |
+
"loss": 1.1218,
|
| 129 |
"step": 1200
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 3.81,
|
|
|
|
| 133 |
"learning_rate": 0.0001,
|
| 134 |
+
"loss": 1.1231,
|
| 135 |
"step": 1300
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 4.0,
|
| 139 |
+
"eval_accuracy": 0.45697435897435895,
|
| 140 |
+
"eval_loss": 3.7999510765075684,
|
| 141 |
+
"eval_runtime": 4.0886,
|
| 142 |
+
"eval_samples_per_second": 122.29,
|
| 143 |
+
"eval_steps_per_second": 15.409,
|
| 144 |
"step": 1366
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 4.0,
|
| 148 |
+
"eval_bleu": 0.15413978034957934,
|
| 149 |
+
"eval_exact_match": 0.0,
|
| 150 |
+
"eval_prefix_exact_match": 0.054,
|
| 151 |
"step": 1366
|
| 152 |
},
|
| 153 |
{
|
| 154 |
+
"epoch": 4.1,
|
|
|
|
| 155 |
"learning_rate": 0.0001,
|
| 156 |
+
"loss": 1.0823,
|
| 157 |
"step": 1400
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 4.39,
|
|
|
|
| 161 |
"learning_rate": 0.0001,
|
| 162 |
+
"loss": 0.997,
|
| 163 |
"step": 1500
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 4.69,
|
|
|
|
| 167 |
"learning_rate": 0.0001,
|
| 168 |
+
"loss": 1.0024,
|
| 169 |
"step": 1600
|
| 170 |
},
|
| 171 |
{
|
| 172 |
+
"epoch": 4.98,
|
|
|
|
| 173 |
"learning_rate": 0.0001,
|
| 174 |
+
"loss": 0.995,
|
| 175 |
"step": 1700
|
| 176 |
},
|
| 177 |
{
|
| 178 |
+
"epoch": 5.0,
|
| 179 |
+
"eval_accuracy": 0.4552307692307692,
|
| 180 |
+
"eval_loss": 3.953216314315796,
|
| 181 |
+
"eval_runtime": 4.9824,
|
| 182 |
+
"eval_samples_per_second": 100.354,
|
| 183 |
+
"eval_steps_per_second": 12.645,
|
| 184 |
"step": 1707
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 5.0,
|
| 188 |
+
"eval_bleu": 0.1611344512465791,
|
| 189 |
+
"eval_exact_match": 0.0,
|
| 190 |
+
"eval_prefix_exact_match": 0.046,
|
| 191 |
"step": 1707
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 5.27,
|
|
|
|
| 195 |
"learning_rate": 0.0001,
|
| 196 |
+
"loss": 0.8495,
|
| 197 |
"step": 1800
|
| 198 |
},
|
| 199 |
{
|
| 200 |
+
"epoch": 5.56,
|
|
|
|
| 201 |
"learning_rate": 0.0001,
|
| 202 |
+
"loss": 0.854,
|
| 203 |
"step": 1900
|
| 204 |
},
|
| 205 |
{
|
| 206 |
+
"epoch": 5.86,
|
|
|
|
| 207 |
"learning_rate": 0.0001,
|
| 208 |
+
"loss": 0.8693,
|
| 209 |
"step": 2000
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 6.0,
|
| 213 |
+
"eval_accuracy": 0.45261538461538464,
|
| 214 |
+
"eval_loss": 4.0766072273254395,
|
| 215 |
+
"eval_runtime": 4.8095,
|
| 216 |
+
"eval_samples_per_second": 103.961,
|
| 217 |
+
"eval_steps_per_second": 13.099,
|
| 218 |
"step": 2049
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 6.0,
|
| 222 |
+
"eval_bleu": 0.1489904300924264,
|
| 223 |
+
"eval_exact_match": 0.0,
|
| 224 |
+
"eval_prefix_exact_match": 0.044,
|
| 225 |
"step": 2049
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 6.15,
|
|
|
|
| 229 |
"learning_rate": 0.0001,
|
| 230 |
+
"loss": 0.7917,
|
| 231 |
"step": 2100
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 6.44,
|
|
|
|
| 235 |
"learning_rate": 0.0001,
|
| 236 |
+
"loss": 0.7061,
|
| 237 |
"step": 2200
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 6.73,
|
|
|
|
| 241 |
"learning_rate": 0.0001,
|
| 242 |
+
"loss": 0.7302,
|
| 243 |
"step": 2300
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 7.0,
|
| 247 |
+
"eval_accuracy": 0.4501025641025641,
|
| 248 |
+
"eval_loss": 4.171727180480957,
|
| 249 |
+
"eval_runtime": 4.2169,
|
| 250 |
+
"eval_samples_per_second": 118.57,
|
| 251 |
+
"eval_steps_per_second": 14.94,
|
| 252 |
"step": 2390
|
| 253 |
},
|
| 254 |
{
|
| 255 |
+
"epoch": 7.0,
|
|
|
|
|
|
|
|
|
|
| 256 |
"step": 2390
|
| 257 |
},
|
| 258 |
{
|
| 259 |
+
"epoch": 7.03,
|
|
|
|
| 260 |
"learning_rate": 0.0001,
|
| 261 |
+
"loss": 0.7227,
|
| 262 |
"step": 2400
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 7.32,
|
|
|
|
| 266 |
"learning_rate": 0.0001,
|
| 267 |
+
"loss": 0.573,
|
| 268 |
"step": 2500
|
| 269 |
},
|
| 270 |
{
|
| 271 |
+
"epoch": 7.61,
|
|
|
|
| 272 |
"learning_rate": 0.0001,
|
| 273 |
+
"loss": 0.6036,
|
| 274 |
"step": 2600
|
| 275 |
},
|
| 276 |
{
|
| 277 |
+
"epoch": 7.91,
|
|
|
|
| 278 |
"learning_rate": 0.0001,
|
| 279 |
+
"loss": 0.6033,
|
| 280 |
"step": 2700
|
| 281 |
},
|
| 282 |
{
|
| 283 |
"epoch": 8.0,
|
| 284 |
+
"eval_accuracy": 0.448,
|
| 285 |
+
"eval_loss": 4.277780055999756,
|
| 286 |
+
"eval_runtime": 4.0947,
|
| 287 |
+
"eval_samples_per_second": 122.11,
|
| 288 |
+
"eval_steps_per_second": 15.386,
|
| 289 |
"step": 2732
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 8.0,
|
|
|
|
|
|
|
|
|
|
| 293 |
"step": 2732
|
| 294 |
},
|
| 295 |
{
|
| 296 |
+
"epoch": 8.2,
|
|
|
|
| 297 |
"learning_rate": 0.0001,
|
| 298 |
+
"loss": 0.4945,
|
| 299 |
"step": 2800
|
| 300 |
},
|
| 301 |
{
|
| 302 |
+
"epoch": 8.49,
|
|
|
|
| 303 |
"learning_rate": 0.0001,
|
| 304 |
+
"loss": 0.4718,
|
| 305 |
"step": 2900
|
| 306 |
},
|
| 307 |
{
|
| 308 |
+
"epoch": 8.78,
|
|
|
|
| 309 |
"learning_rate": 0.0001,
|
| 310 |
+
"loss": 0.4825,
|
| 311 |
"step": 3000
|
| 312 |
},
|
| 313 |
{
|
| 314 |
+
"epoch": 9.0,
|
| 315 |
+
"eval_accuracy": 0.4462051282051282,
|
| 316 |
+
"eval_loss": 4.34151554107666,
|
| 317 |
+
"eval_runtime": 4.8113,
|
| 318 |
+
"eval_samples_per_second": 103.921,
|
| 319 |
+
"eval_steps_per_second": 13.094,
|
| 320 |
"step": 3073
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"epoch": 9.0,
|
|
|
|
|
|
|
|
|
|
| 324 |
"step": 3073
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 9.08,
|
|
|
|
| 328 |
"learning_rate": 0.0001,
|
| 329 |
+
"loss": 0.4538,
|
| 330 |
"step": 3100
|
| 331 |
},
|
| 332 |
{
|
| 333 |
+
"epoch": 9.37,
|
|
|
|
| 334 |
"learning_rate": 0.0001,
|
| 335 |
+
"loss": 0.3609,
|
| 336 |
"step": 3200
|
| 337 |
},
|
| 338 |
{
|
| 339 |
+
"epoch": 9.66,
|
|
|
|
| 340 |
"learning_rate": 0.0001,
|
| 341 |
+
"loss": 0.3769,
|
| 342 |
"step": 3300
|
| 343 |
},
|
| 344 |
{
|
| 345 |
+
"epoch": 9.96,
|
|
|
|
| 346 |
"learning_rate": 0.0001,
|
| 347 |
+
"loss": 0.387,
|
| 348 |
"step": 3400
|
| 349 |
},
|
| 350 |
{
|
| 351 |
"epoch": 10.0,
|
| 352 |
+
"eval_accuracy": 0.4463076923076923,
|
| 353 |
+
"eval_loss": 4.413115978240967,
|
| 354 |
+
"eval_runtime": 4.2874,
|
| 355 |
+
"eval_samples_per_second": 116.62,
|
| 356 |
+
"eval_steps_per_second": 14.694,
|
| 357 |
"step": 3415
|
| 358 |
},
|
| 359 |
{
|
| 360 |
"epoch": 10.0,
|
|
|
|
|
|
|
|
|
|
| 361 |
"step": 3415
|
| 362 |
},
|
| 363 |
{
|
| 364 |
+
"epoch": 10.25,
|
|
|
|
| 365 |
"learning_rate": 0.0001,
|
| 366 |
+
"loss": 0.2911,
|
| 367 |
"step": 3500
|
| 368 |
},
|
| 369 |
{
|
| 370 |
+
"epoch": 10.54,
|
|
|
|
| 371 |
"learning_rate": 0.0001,
|
| 372 |
+
"loss": 0.2881,
|
| 373 |
"step": 3600
|
| 374 |
},
|
| 375 |
{
|
| 376 |
+
"epoch": 10.83,
|
|
|
|
| 377 |
"learning_rate": 0.0001,
|
| 378 |
+
"loss": 0.2933,
|
| 379 |
"step": 3700
|
| 380 |
},
|
| 381 |
{
|
| 382 |
+
"epoch": 11.0,
|
| 383 |
+
"eval_accuracy": 0.44343589743589745,
|
| 384 |
+
"eval_loss": 4.490556716918945,
|
| 385 |
+
"eval_runtime": 4.2824,
|
| 386 |
+
"eval_samples_per_second": 116.757,
|
| 387 |
+
"eval_steps_per_second": 14.711,
|
| 388 |
"step": 3756
|
| 389 |
},
|
| 390 |
{
|
| 391 |
+
"epoch": 11.0,
|
|
|
|
|
|
|
|
|
|
| 392 |
"step": 3756
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"epoch": 11.13,
|
|
|
|
| 396 |
"learning_rate": 0.0001,
|
| 397 |
+
"loss": 0.2653,
|
| 398 |
"step": 3800
|
| 399 |
},
|
| 400 |
{
|
| 401 |
+
"epoch": 11.42,
|
|
|
|
| 402 |
"learning_rate": 0.0001,
|
| 403 |
+
"loss": 0.2181,
|
| 404 |
"step": 3900
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 11.71,
|
|
|
|
| 408 |
"learning_rate": 0.0001,
|
| 409 |
+
"loss": 0.2344,
|
| 410 |
"step": 4000
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 12.0,
|
| 414 |
+
"eval_accuracy": 0.44251282051282054,
|
| 415 |
+
"eval_loss": 4.651719570159912,
|
| 416 |
+
"eval_runtime": 4.0941,
|
| 417 |
+
"eval_samples_per_second": 122.126,
|
| 418 |
+
"eval_steps_per_second": 15.388,
|
| 419 |
"step": 4098
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 12.0,
|
|
|
|
|
|
|
|
|
|
| 423 |
"step": 4098
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 12.01,
|
|
|
|
| 427 |
"learning_rate": 0.0001,
|
| 428 |
+
"loss": 0.2414,
|
| 429 |
"step": 4100
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 12.3,
|
|
|
|
| 433 |
"learning_rate": 0.0001,
|
| 434 |
+
"loss": 0.173,
|
| 435 |
"step": 4200
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 12.59,
|
|
|
|
| 439 |
"learning_rate": 0.0001,
|
| 440 |
+
"loss": 0.183,
|
| 441 |
"step": 4300
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 12.88,
|
|
|
|
| 445 |
"learning_rate": 0.0001,
|
| 446 |
+
"loss": 0.1919,
|
| 447 |
"step": 4400
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 13.0,
|
| 451 |
+
"eval_accuracy": 0.44076923076923075,
|
| 452 |
+
"eval_loss": 4.751525402069092,
|
| 453 |
+
"eval_runtime": 4.0902,
|
| 454 |
+
"eval_samples_per_second": 122.244,
|
| 455 |
+
"eval_steps_per_second": 15.403,
|
| 456 |
"step": 4439
|
| 457 |
},
|
| 458 |
{
|
| 459 |
+
"epoch": 13.0,
|
|
|
|
|
|
|
|
|
|
| 460 |
"step": 4439
|
| 461 |
},
|
| 462 |
{
|
| 463 |
+
"epoch": 13.18,
|
|
|
|
| 464 |
"learning_rate": 0.0001,
|
| 465 |
+
"loss": 0.1661,
|
| 466 |
"step": 4500
|
| 467 |
},
|
| 468 |
{
|
| 469 |
+
"epoch": 13.47,
|
|
|
|
| 470 |
"learning_rate": 0.0001,
|
| 471 |
+
"loss": 0.1501,
|
| 472 |
"step": 4600
|
| 473 |
},
|
| 474 |
{
|
| 475 |
+
"epoch": 13.76,
|
|
|
|
| 476 |
"learning_rate": 0.0001,
|
| 477 |
+
"loss": 0.1581,
|
| 478 |
"step": 4700
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 14.0,
|
| 482 |
+
"eval_accuracy": 0.442051282051282,
|
| 483 |
+
"eval_loss": 4.7323455810546875,
|
| 484 |
+
"eval_runtime": 4.8128,
|
| 485 |
+
"eval_samples_per_second": 103.89,
|
| 486 |
+
"eval_steps_per_second": 13.09,
|
| 487 |
"step": 4781
|
| 488 |
},
|
| 489 |
{
|
| 490 |
"epoch": 14.0,
|
| 491 |
+
"eval_bleu": 0.01991302915032707,
|
| 492 |
+
"eval_exact_match": 0.0,
|
| 493 |
+
"eval_prefix_exact_match": 0.0,
|
| 494 |
"step": 4781
|
| 495 |
},
|
| 496 |
{
|
| 497 |
+
"epoch": 14.06,
|
|
|
|
| 498 |
"learning_rate": 0.0001,
|
| 499 |
+
"loss": 0.1563,
|
| 500 |
"step": 4800
|
| 501 |
},
|
| 502 |
{
|
| 503 |
+
"epoch": 14.35,
|
|
|
|
| 504 |
"learning_rate": 0.0001,
|
| 505 |
+
"loss": 0.13,
|
| 506 |
"step": 4900
|
| 507 |
},
|
| 508 |
{
|
| 509 |
+
"epoch": 14.64,
|
|
|
|
| 510 |
"learning_rate": 0.0001,
|
| 511 |
+
"loss": 0.1388,
|
| 512 |
"step": 5000
|
| 513 |
},
|
| 514 |
{
|
| 515 |
+
"epoch": 14.93,
|
|
|
|
| 516 |
"learning_rate": 0.0001,
|
| 517 |
+
"loss": 0.1429,
|
| 518 |
"step": 5100
|
| 519 |
},
|
| 520 |
{
|
| 521 |
+
"epoch": 15.0,
|
| 522 |
+
"eval_accuracy": 0.44066666666666665,
|
| 523 |
+
"eval_loss": 4.810059547424316,
|
| 524 |
+
"eval_runtime": 5.0086,
|
| 525 |
+
"eval_samples_per_second": 99.829,
|
| 526 |
+
"eval_steps_per_second": 12.578,
|
| 527 |
"step": 5122
|
| 528 |
},
|
| 529 |
{
|
| 530 |
+
"epoch": 15.0,
|
| 531 |
+
"eval_bleu": 0.007355658423274405,
|
| 532 |
"eval_exact_match": 0.0,
|
| 533 |
+
"eval_prefix_exact_match": 0.002,
|
| 534 |
"step": 5122
|
| 535 |
},
|
| 536 |
{
|
| 537 |
+
"epoch": 15.23,
|
|
|
|
| 538 |
"learning_rate": 0.0001,
|
| 539 |
+
"loss": 0.1228,
|
| 540 |
"step": 5200
|
| 541 |
},
|
| 542 |
{
|
| 543 |
+
"epoch": 15.52,
|
|
|
|
| 544 |
"learning_rate": 0.0001,
|
| 545 |
+
"loss": 0.1233,
|
| 546 |
"step": 5300
|
| 547 |
},
|
| 548 |
{
|
| 549 |
+
"epoch": 15.81,
|
|
|
|
| 550 |
"learning_rate": 0.0001,
|
| 551 |
+
"loss": 0.1279,
|
| 552 |
"step": 5400
|
| 553 |
},
|
| 554 |
{
|
| 555 |
"epoch": 16.0,
|
| 556 |
+
"eval_accuracy": 0.44061538461538463,
|
| 557 |
+
"eval_loss": 4.84822940826416,
|
| 558 |
+
"eval_runtime": 4.1962,
|
| 559 |
+
"eval_samples_per_second": 119.156,
|
| 560 |
+
"eval_steps_per_second": 15.014,
|
| 561 |
"step": 5464
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 16.0,
|
| 565 |
+
"eval_bleu": 0.020518239891916933,
|
| 566 |
+
"eval_exact_match": 0.0,
|
| 567 |
+
"eval_prefix_exact_match": 0.0,
|
| 568 |
"step": 5464
|
| 569 |
},
|
| 570 |
{
|
| 571 |
+
"epoch": 16.11,
|
|
|
|
| 572 |
"learning_rate": 0.0001,
|
| 573 |
+
"loss": 0.1248,
|
| 574 |
"step": 5500
|
| 575 |
},
|
| 576 |
{
|
| 577 |
+
"epoch": 16.4,
|
|
|
|
| 578 |
"learning_rate": 0.0001,
|
| 579 |
+
"loss": 0.1118,
|
| 580 |
"step": 5600
|
| 581 |
},
|
| 582 |
{
|
| 583 |
+
"epoch": 16.69,
|
|
|
|
| 584 |
"learning_rate": 0.0001,
|
| 585 |
+
"loss": 0.1172,
|
| 586 |
"step": 5700
|
| 587 |
},
|
| 588 |
{
|
| 589 |
+
"epoch": 16.98,
|
|
|
|
| 590 |
"learning_rate": 0.0001,
|
| 591 |
+
"loss": 0.1231,
|
| 592 |
"step": 5800
|
| 593 |
},
|
| 594 |
{
|
| 595 |
+
"epoch": 17.0,
|
| 596 |
+
"eval_accuracy": 0.4410769230769231,
|
| 597 |
+
"eval_loss": 4.973475933074951,
|
| 598 |
+
"eval_runtime": 4.4042,
|
| 599 |
+
"eval_samples_per_second": 113.529,
|
| 600 |
+
"eval_steps_per_second": 14.305,
|
| 601 |
"step": 5805
|
| 602 |
},
|
| 603 |
{
|
| 604 |
+
"epoch": 17.0,
|
| 605 |
+
"eval_bleu": 0.025070041424787165,
|
| 606 |
"eval_exact_match": 0.0,
|
| 607 |
+
"eval_prefix_exact_match": 0.0,
|
| 608 |
"step": 5805
|
| 609 |
},
|
| 610 |
{
|
| 611 |
+
"epoch": 17.28,
|
|
|
|
| 612 |
"learning_rate": 0.0001,
|
| 613 |
+
"loss": 0.1068,
|
| 614 |
"step": 5900
|
| 615 |
},
|
| 616 |
{
|
| 617 |
+
"epoch": 17.57,
|
|
|
|
| 618 |
"learning_rate": 0.0001,
|
| 619 |
+
"loss": 0.111,
|
| 620 |
"step": 6000
|
| 621 |
},
|
| 622 |
{
|
| 623 |
+
"epoch": 17.86,
|
|
|
|
| 624 |
"learning_rate": 0.0001,
|
| 625 |
+
"loss": 0.1145,
|
| 626 |
"step": 6100
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 18.0,
|
| 630 |
+
"eval_accuracy": 0.44153846153846155,
|
| 631 |
+
"eval_loss": 5.012055397033691,
|
| 632 |
+
"eval_runtime": 4.1955,
|
| 633 |
+
"eval_samples_per_second": 119.174,
|
| 634 |
+
"eval_steps_per_second": 15.016,
|
| 635 |
"step": 6147
|
| 636 |
},
|
| 637 |
{
|
| 638 |
"epoch": 18.0,
|
| 639 |
+
"eval_bleu": 0.0018720973379033757,
|
| 640 |
"eval_exact_match": 0.0,
|
| 641 |
+
"eval_prefix_exact_match": 0.0,
|
| 642 |
"step": 6147
|
| 643 |
},
|
| 644 |
{
|
| 645 |
+
"epoch": 18.16,
|
|
|
|
| 646 |
"learning_rate": 0.0001,
|
| 647 |
+
"loss": 0.1093,
|
| 648 |
"step": 6200
|
| 649 |
},
|
| 650 |
{
|
| 651 |
+
"epoch": 18.45,
|
|
|
|
| 652 |
"learning_rate": 0.0001,
|
| 653 |
+
"loss": 0.1051,
|
| 654 |
"step": 6300
|
| 655 |
},
|
| 656 |
{
|
| 657 |
+
"epoch": 18.74,
|
|
|
|
| 658 |
"learning_rate": 0.0001,
|
| 659 |
+
"loss": 0.1087,
|
| 660 |
"step": 6400
|
| 661 |
},
|
| 662 |
{
|
| 663 |
+
"epoch": 19.0,
|
| 664 |
+
"eval_accuracy": 0.43943589743589745,
|
| 665 |
+
"eval_loss": 4.983563423156738,
|
| 666 |
+
"eval_runtime": 4.0932,
|
| 667 |
+
"eval_samples_per_second": 122.154,
|
| 668 |
+
"eval_steps_per_second": 15.391,
|
| 669 |
"step": 6488
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 19.0,
|
| 673 |
+
"eval_bleu": 0.0014273377715875865,
|
| 674 |
"eval_exact_match": 0.0,
|
| 675 |
+
"eval_prefix_exact_match": 0.0,
|
| 676 |
"step": 6488
|
| 677 |
},
|
| 678 |
{
|
| 679 |
+
"epoch": 19.03,
|
|
|
|
| 680 |
"learning_rate": 0.0001,
|
| 681 |
+
"loss": 0.1115,
|
| 682 |
"step": 6500
|
| 683 |
},
|
| 684 |
{
|
| 685 |
+
"epoch": 19.33,
|
|
|
|
| 686 |
"learning_rate": 0.0001,
|
| 687 |
+
"loss": 0.1001,
|
| 688 |
"step": 6600
|
| 689 |
},
|
| 690 |
{
|
| 691 |
+
"epoch": 19.62,
|
|
|
|
| 692 |
"learning_rate": 0.0001,
|
| 693 |
+
"loss": 0.1049,
|
| 694 |
"step": 6700
|
| 695 |
},
|
| 696 |
{
|
| 697 |
+
"epoch": 19.91,
|
|
|
|
| 698 |
"learning_rate": 0.0001,
|
| 699 |
+
"loss": 0.1084,
|
| 700 |
"step": 6800
|
| 701 |
},
|
| 702 |
{
|
| 703 |
"epoch": 20.0,
|
| 704 |
+
"eval_accuracy": 0.4388205128205128,
|
| 705 |
+
"eval_loss": 5.117064952850342,
|
| 706 |
+
"eval_runtime": 4.0908,
|
| 707 |
+
"eval_samples_per_second": 122.227,
|
| 708 |
+
"eval_steps_per_second": 15.401,
|
| 709 |
"step": 6830
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 20.0,
|
| 713 |
+
"eval_bleu": 0.004914730802266865,
|
| 714 |
"eval_exact_match": 0.0,
|
| 715 |
+
"eval_prefix_exact_match": 0.0,
|
| 716 |
"step": 6830
|
| 717 |
},
|
| 718 |
{
|
| 719 |
+
"epoch": 20.2,
|
|
|
|
| 720 |
"learning_rate": 0.0001,
|
| 721 |
+
"loss": 0.102,
|
| 722 |
"step": 6900
|
| 723 |
},
|
| 724 |
{
|
| 725 |
+
"epoch": 20.5,
|
|
|
|
| 726 |
"learning_rate": 0.0001,
|
| 727 |
+
"loss": 0.1014,
|
| 728 |
"step": 7000
|
| 729 |
},
|
| 730 |
{
|
| 731 |
+
"epoch": 20.79,
|
|
|
|
| 732 |
"learning_rate": 0.0001,
|
| 733 |
+
"loss": 0.1069,
|
| 734 |
"step": 7100
|
| 735 |
},
|
| 736 |
{
|
| 737 |
+
"epoch": 21.0,
|
| 738 |
+
"eval_accuracy": 0.44051282051282054,
|
| 739 |
+
"eval_loss": 5.011991500854492,
|
| 740 |
+
"eval_runtime": 4.1961,
|
| 741 |
+
"eval_samples_per_second": 119.159,
|
| 742 |
+
"eval_steps_per_second": 15.014,
|
| 743 |
"step": 7171
|
| 744 |
},
|
| 745 |
{
|
| 746 |
+
"epoch": 21.0,
|
|
|
|
|
|
|
|
|
|
| 747 |
"step": 7171
|
| 748 |
},
|
| 749 |
{
|
| 750 |
+
"epoch": 21.08,
|
|
|
|
| 751 |
"learning_rate": 0.0001,
|
| 752 |
+
"loss": 0.1075,
|
| 753 |
"step": 7200
|
| 754 |
},
|
| 755 |
{
|
| 756 |
+
"epoch": 21.38,
|
|
|
|
| 757 |
"learning_rate": 0.0001,
|
| 758 |
+
"loss": 0.0985,
|
| 759 |
"step": 7300
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 21.67,
|
|
|
|
| 763 |
"learning_rate": 0.0001,
|
| 764 |
+
"loss": 0.1032,
|
| 765 |
"step": 7400
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 21.96,
|
|
|
|
| 769 |
"learning_rate": 0.0001,
|
| 770 |
+
"loss": 0.1075,
|
| 771 |
"step": 7500
|
| 772 |
},
|
| 773 |
{
|
| 774 |
"epoch": 22.0,
|
| 775 |
+
"eval_accuracy": 0.44,
|
| 776 |
+
"eval_loss": 5.234306812286377,
|
| 777 |
+
"eval_runtime": 4.4185,
|
| 778 |
+
"eval_samples_per_second": 113.162,
|
| 779 |
+
"eval_steps_per_second": 14.258,
|
| 780 |
"step": 7513
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 22.0,
|
|
|
|
|
|
|
|
|
|
| 784 |
"step": 7513
|
| 785 |
},
|
| 786 |
{
|
| 787 |
+
"epoch": 22.25,
|
|
|
|
| 788 |
"learning_rate": 0.0001,
|
| 789 |
+
"loss": 0.0959,
|
| 790 |
"step": 7600
|
| 791 |
},
|
| 792 |
{
|
| 793 |
+
"epoch": 22.55,
|
|
|
|
| 794 |
"learning_rate": 0.0001,
|
| 795 |
+
"loss": 0.0998,
|
| 796 |
"step": 7700
|
| 797 |
},
|
| 798 |
{
|
| 799 |
+
"epoch": 22.84,
|
|
|
|
| 800 |
"learning_rate": 0.0001,
|
| 801 |
+
"loss": 0.1024,
|
| 802 |
"step": 7800
|
| 803 |
},
|
| 804 |
{
|
| 805 |
+
"epoch": 23.0,
|
| 806 |
+
"eval_accuracy": 0.4408717948717949,
|
| 807 |
+
"eval_loss": 5.150094985961914,
|
| 808 |
+
"eval_runtime": 4.2006,
|
| 809 |
+
"eval_samples_per_second": 119.032,
|
| 810 |
+
"eval_steps_per_second": 14.998,
|
| 811 |
"step": 7854
|
| 812 |
},
|
| 813 |
{
|
| 814 |
+
"epoch": 23.0,
|
| 815 |
+
"eval_bleu": 0.008507706871345696,
|
| 816 |
+
"eval_exact_match": 0.0,
|
| 817 |
+
"eval_prefix_exact_match": 0.002,
|
| 818 |
"step": 7854
|
| 819 |
},
|
| 820 |
{
|
| 821 |
+
"epoch": 23.13,
|
|
|
|
| 822 |
"learning_rate": 0.0001,
|
| 823 |
+
"loss": 0.0997,
|
| 824 |
"step": 7900
|
| 825 |
},
|
| 826 |
{
|
| 827 |
+
"epoch": 23.43,
|
|
|
|
| 828 |
"learning_rate": 0.0001,
|
| 829 |
+
"loss": 0.0946,
|
| 830 |
"step": 8000
|
| 831 |
},
|
| 832 |
{
|
| 833 |
+
"epoch": 23.72,
|
|
|
|
| 834 |
"learning_rate": 0.0001,
|
| 835 |
+
"loss": 0.0981,
|
| 836 |
"step": 8100
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 24.0,
|
| 840 |
+
"eval_accuracy": 0.44025641025641027,
|
| 841 |
+
"eval_loss": 5.080117225646973,
|
| 842 |
+
"eval_runtime": 4.1942,
|
| 843 |
+
"eval_samples_per_second": 119.212,
|
| 844 |
+
"eval_steps_per_second": 15.021,
|
| 845 |
"step": 8196
|
| 846 |
},
|
| 847 |
{
|
| 848 |
"epoch": 24.0,
|
|
|
|
|
|
|
|
|
|
| 849 |
"step": 8196
|
| 850 |
},
|
| 851 |
{
|
| 852 |
+
"epoch": 24.01,
|
|
|
|
| 853 |
"learning_rate": 0.0001,
|
| 854 |
+
"loss": 0.101,
|
| 855 |
"step": 8200
|
| 856 |
},
|
| 857 |
{
|
| 858 |
+
"epoch": 24.3,
|
|
|
|
| 859 |
"learning_rate": 0.0001,
|
| 860 |
+
"loss": 0.0923,
|
| 861 |
"step": 8300
|
| 862 |
},
|
| 863 |
{
|
| 864 |
+
"epoch": 24.6,
|
|
|
|
| 865 |
"learning_rate": 0.0001,
|
| 866 |
+
"loss": 0.095,
|
| 867 |
"step": 8400
|
| 868 |
},
|
| 869 |
{
|
| 870 |
+
"epoch": 24.89,
|
|
|
|
| 871 |
"learning_rate": 0.0001,
|
| 872 |
+
"loss": 0.097,
|
| 873 |
"step": 8500
|
| 874 |
},
|
| 875 |
{
|
| 876 |
+
"epoch": 25.0,
|
| 877 |
+
"eval_accuracy": 0.44158974358974357,
|
| 878 |
+
"eval_loss": 5.103724956512451,
|
| 879 |
+
"eval_runtime": 4.307,
|
| 880 |
+
"eval_samples_per_second": 116.09,
|
| 881 |
+
"eval_steps_per_second": 14.627,
|
| 882 |
"step": 8537
|
| 883 |
},
|
| 884 |
{
|
| 885 |
+
"epoch": 25.0,
|
|
|
|
|
|
|
|
|
|
| 886 |
"step": 8537
|
| 887 |
},
|
| 888 |
{
|
| 889 |
+
"epoch": 25.18,
|
|
|
|
| 890 |
"learning_rate": 0.0001,
|
| 891 |
+
"loss": 0.0925,
|
| 892 |
"step": 8600
|
| 893 |
},
|
| 894 |
{
|
| 895 |
+
"epoch": 25.48,
|
|
|
|
| 896 |
"learning_rate": 0.0001,
|
| 897 |
+
"loss": 0.0926,
|
| 898 |
"step": 8700
|
| 899 |
},
|
| 900 |
{
|
| 901 |
+
"epoch": 25.77,
|
|
|
|
| 902 |
"learning_rate": 0.0001,
|
| 903 |
+
"loss": 0.0963,
|
| 904 |
"step": 8800
|
| 905 |
},
|
| 906 |
{
|
| 907 |
"epoch": 26.0,
|
| 908 |
+
"eval_accuracy": 0.4397948717948718,
|
| 909 |
+
"eval_loss": 5.206360340118408,
|
| 910 |
+
"eval_runtime": 4.8198,
|
| 911 |
+
"eval_samples_per_second": 103.739,
|
| 912 |
+
"eval_steps_per_second": 13.071,
|
| 913 |
"step": 8879
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 26.0,
|
|
|
|
|
|
|
|
|
|
| 917 |
"step": 8879
|
| 918 |
},
|
| 919 |
{
|
| 920 |
+
"epoch": 26.06,
|
|
|
|
| 921 |
"learning_rate": 0.0001,
|
| 922 |
+
"loss": 0.0966,
|
| 923 |
"step": 8900
|
| 924 |
},
|
| 925 |
{
|
| 926 |
+
"epoch": 26.35,
|
|
|
|
| 927 |
"learning_rate": 0.0001,
|
| 928 |
+
"loss": 0.0909,
|
| 929 |
"step": 9000
|
| 930 |
},
|
| 931 |
{
|
| 932 |
+
"epoch": 26.65,
|
|
|
|
| 933 |
"learning_rate": 0.0001,
|
| 934 |
+
"loss": 0.0951,
|
| 935 |
"step": 9100
|
| 936 |
},
|
| 937 |
{
|
| 938 |
+
"epoch": 26.94,
|
|
|
|
| 939 |
"learning_rate": 0.0001,
|
| 940 |
+
"loss": 0.0983,
|
| 941 |
"step": 9200
|
| 942 |
},
|
| 943 |
{
|
| 944 |
+
"epoch": 27.0,
|
| 945 |
+
"eval_accuracy": 0.4413846153846154,
|
| 946 |
+
"eval_loss": 5.066434383392334,
|
| 947 |
+
"eval_runtime": 4.1912,
|
| 948 |
+
"eval_samples_per_second": 119.298,
|
| 949 |
+
"eval_steps_per_second": 15.032,
|
| 950 |
"step": 9220
|
| 951 |
},
|
| 952 |
{
|
| 953 |
+
"epoch": 27.0,
|
|
|
|
|
|
|
|
|
|
| 954 |
"step": 9220
|
| 955 |
},
|
| 956 |
{
|
| 957 |
+
"epoch": 27.23,
|
|
|
|
| 958 |
"learning_rate": 0.0001,
|
| 959 |
+
"loss": 0.092,
|
| 960 |
"step": 9300
|
| 961 |
},
|
| 962 |
{
|
| 963 |
+
"epoch": 27.53,
|
|
|
|
| 964 |
"learning_rate": 0.0001,
|
| 965 |
+
"loss": 0.0935,
|
| 966 |
"step": 9400
|
| 967 |
},
|
| 968 |
{
|
| 969 |
+
"epoch": 27.82,
|
|
|
|
| 970 |
"learning_rate": 0.0001,
|
| 971 |
+
"loss": 0.0969,
|
| 972 |
"step": 9500
|
| 973 |
},
|
| 974 |
{
|
| 975 |
"epoch": 28.0,
|
| 976 |
+
"eval_accuracy": 0.440974358974359,
|
| 977 |
+
"eval_loss": 5.255864143371582,
|
| 978 |
+
"eval_runtime": 4.1948,
|
| 979 |
+
"eval_samples_per_second": 119.194,
|
| 980 |
+
"eval_steps_per_second": 15.018,
|
| 981 |
"step": 9562
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 28.0,
|
|
|
|
|
|
|
|
|
|
| 985 |
"step": 9562
|
| 986 |
},
|
| 987 |
{
|
| 988 |
+
"epoch": 28.11,
|
|
|
|
| 989 |
"learning_rate": 0.0001,
|
| 990 |
+
"loss": 0.0947,
|
| 991 |
"step": 9600
|
| 992 |
},
|
| 993 |
{
|
| 994 |
+
"epoch": 28.4,
|
|
|
|
| 995 |
"learning_rate": 0.0001,
|
| 996 |
+
"loss": 0.0915,
|
| 997 |
"step": 9700
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
+
"epoch": 28.7,
|
|
|
|
| 1001 |
"learning_rate": 0.0001,
|
| 1002 |
+
"loss": 0.0948,
|
| 1003 |
"step": 9800
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
+
"epoch": 28.99,
|
|
|
|
| 1007 |
"learning_rate": 0.0001,
|
| 1008 |
+
"loss": 0.0966,
|
| 1009 |
"step": 9900
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
+
"epoch": 29.0,
|
| 1013 |
+
"eval_accuracy": 0.44035897435897436,
|
| 1014 |
+
"eval_loss": 5.196049690246582,
|
| 1015 |
+
"eval_runtime": 5.1571,
|
| 1016 |
+
"eval_samples_per_second": 96.953,
|
| 1017 |
+
"eval_steps_per_second": 12.216,
|
| 1018 |
"step": 9903
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
+
"epoch": 29.0,
|
|
|
|
|
|
|
|
|
|
| 1022 |
"step": 9903
|
| 1023 |
},
|
| 1024 |
{
|
| 1025 |
+
"epoch": 29.28,
|
|
|
|
| 1026 |
"learning_rate": 0.0001,
|
| 1027 |
+
"loss": 0.0884,
|
| 1028 |
"step": 10000
|
| 1029 |
},
|
| 1030 |
{
|
| 1031 |
+
"epoch": 29.58,
|
|
|
|
| 1032 |
"learning_rate": 0.0001,
|
| 1033 |
+
"loss": 0.0913,
|
| 1034 |
"step": 10100
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
+
"epoch": 29.87,
|
|
|
|
| 1038 |
"learning_rate": 0.0001,
|
| 1039 |
+
"loss": 0.0954,
|
| 1040 |
"step": 10200
|
| 1041 |
},
|
| 1042 |
{
|
| 1043 |
"epoch": 30.0,
|
| 1044 |
+
"eval_accuracy": 0.43958974358974356,
|
| 1045 |
+
"eval_loss": 5.223780632019043,
|
| 1046 |
+
"eval_runtime": 4.8025,
|
| 1047 |
+
"eval_samples_per_second": 104.112,
|
| 1048 |
+
"eval_steps_per_second": 13.118,
|
| 1049 |
"step": 10245
|
| 1050 |
},
|
| 1051 |
{
|
| 1052 |
"epoch": 30.0,
|
|
|
|
|
|
|
|
|
|
| 1053 |
"step": 10245
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
+
"epoch": 30.16,
|
|
|
|
| 1057 |
"learning_rate": 0.0001,
|
| 1058 |
+
"loss": 0.0913,
|
| 1059 |
"step": 10300
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 30.45,
|
|
|
|
| 1063 |
"learning_rate": 0.0001,
|
| 1064 |
+
"loss": 0.0901,
|
| 1065 |
"step": 10400
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
+
"epoch": 30.75,
|
|
|
|
| 1069 |
"learning_rate": 0.0001,
|
| 1070 |
+
"loss": 0.0931,
|
| 1071 |
"step": 10500
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
+
"epoch": 31.0,
|
| 1075 |
+
"eval_accuracy": 0.4402051282051282,
|
| 1076 |
+
"eval_loss": 5.219459056854248,
|
| 1077 |
+
"eval_runtime": 4.2447,
|
| 1078 |
+
"eval_samples_per_second": 117.794,
|
| 1079 |
+
"eval_steps_per_second": 14.842,
|
| 1080 |
"step": 10586
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
+
"epoch": 31.0,
|
|
|
|
|
|
|
|
|
|
| 1084 |
"step": 10586
|
| 1085 |
}
|
| 1086 |
],
|
| 1087 |
"logging_steps": 100,
|
| 1088 |
"max_steps": 17050,
|
|
|
|
| 1089 |
"num_train_epochs": 50,
|
| 1090 |
"save_steps": 500,
|
| 1091 |
+
"total_flos": 2.0963085334180004e+18,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1092 |
"trial_name": null,
|
| 1093 |
"trial_params": null
|
| 1094 |
}
|
checkpoint-10586/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c60050af374bb2df141f4673c8141d59d83a5bd892d35ffa56ce8af555309ff
|
| 3 |
+
size 4728
|