Andrea Colombo
commited on
Commit
•
1c89313
1
Parent(s):
4793ab9
updated model
Browse files- README.md +1 -4
- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- tokenizer.json +6 -16
- tokenizer_config.json +1 -1
- trainer_state.json +292 -76
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -21,9 +21,6 @@ A Mistral-7B-instruct-v0.1 model to extract a title from the text of Italian law
|
|
| 21 |
- **Finetuned from model:** mistralai/Mistral-7B-Instruct-v0.1
|
| 22 |
|
| 23 |
|
| 24 |
-
## How to Get Started with the Model
|
| 25 |
-
|
| 26 |
-
|
| 27 |
## Training Details
|
| 28 |
|
| 29 |
### Training Procedure
|
|
@@ -34,5 +31,5 @@ We use the paged Adam optimizer, a learning rate of 0.004, and a cosine learning
|
|
| 34 |
|
| 35 |
## Evaluation
|
| 36 |
|
| 37 |
-
The best model reported an evaluation loss of 1.
|
| 38 |
|
|
|
|
| 21 |
- **Finetuned from model:** mistralai/Mistral-7B-Instruct-v0.1
|
| 22 |
|
| 23 |
|
|
|
|
|
|
|
|
|
|
| 24 |
## Training Details
|
| 25 |
|
| 26 |
### Training Procedure
|
|
|
|
| 31 |
|
| 32 |
## Evaluation
|
| 33 |
|
| 34 |
+
The best model reported an evaluation loss of 1.0030452013015747
|
| 35 |
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 109069176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:437dddfa09ec4b75291a5a4b67d5e895271aabe35ccd9f94f8d7cd2f0026b8ce
|
| 3 |
size 109069176
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 218211962
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69c244805725ec0e2048087095e20b54f816a06e887a2c967137a7874a5cddf3
|
| 3 |
size 218211962
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2273825db7916378144c89e647988b4652a7fa2205b00f782cfe68d02a3a468b
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d082bc4e6445151e7041dd6814dd33a4ee1bada774f792aa86b6581d4dd20f7
|
| 3 |
size 1064
|
tokenizer.json
CHANGED
|
@@ -31,23 +31,13 @@
|
|
| 31 |
"special": true
|
| 32 |
}
|
| 33 |
],
|
| 34 |
-
"normalizer":
|
| 35 |
-
|
| 36 |
-
"
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
},
|
| 41 |
-
{
|
| 42 |
-
"type": "Replace",
|
| 43 |
-
"pattern": {
|
| 44 |
-
"String": " "
|
| 45 |
-
},
|
| 46 |
-
"content": "▁"
|
| 47 |
-
}
|
| 48 |
-
]
|
| 49 |
},
|
| 50 |
-
"pre_tokenizer": null,
|
| 51 |
"post_processor": {
|
| 52 |
"type": "TemplateProcessing",
|
| 53 |
"single": [
|
|
|
|
| 31 |
"special": true
|
| 32 |
}
|
| 33 |
],
|
| 34 |
+
"normalizer": null,
|
| 35 |
+
"pre_tokenizer": {
|
| 36 |
+
"type": "Metaspace",
|
| 37 |
+
"replacement": "▁",
|
| 38 |
+
"prepend_scheme": "first",
|
| 39 |
+
"split": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
},
|
|
|
|
| 41 |
"post_processor": {
|
| 42 |
"type": "TemplateProcessing",
|
| 43 |
"single": [
|
tokenizer_config.json
CHANGED
|
@@ -31,7 +31,7 @@
|
|
| 31 |
"bos_token": "<s>",
|
| 32 |
"clean_up_tokenization_spaces": false,
|
| 33 |
"eos_token": "</s>",
|
| 34 |
-
"legacy":
|
| 35 |
"model_max_length": 1000000000000000019884624838656,
|
| 36 |
"pad_token": "</s>",
|
| 37 |
"sp_model_kwargs": {},
|
|
|
|
| 31 |
"bos_token": "<s>",
|
| 32 |
"clean_up_tokenization_spaces": false,
|
| 33 |
"eos_token": "</s>",
|
| 34 |
+
"legacy": false,
|
| 35 |
"model_max_length": 1000000000000000019884624838656,
|
| 36 |
"pad_token": "</s>",
|
| 37 |
"sp_model_kwargs": {},
|
trainer_state.json
CHANGED
|
@@ -1,128 +1,344 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0
|
| 5 |
-
"eval_steps":
|
| 6 |
-
"global_step":
|
| 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.
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
"learning_rate": 0.0002,
|
| 15 |
-
"loss": 1.
|
| 16 |
-
"step":
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
"learning_rate": 0.0002,
|
| 22 |
-
"loss": 1.
|
| 23 |
-
"step":
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"
|
| 28 |
-
"
|
| 29 |
-
"
|
| 30 |
-
"
|
| 31 |
-
"step": 20
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm": 0.
|
| 36 |
"learning_rate": 0.0002,
|
| 37 |
-
"loss": 1.
|
| 38 |
-
"step":
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm": 0.
|
| 43 |
"learning_rate": 0.0002,
|
| 44 |
-
"loss": 1.
|
| 45 |
-
"step":
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"
|
| 50 |
-
"
|
| 51 |
-
"
|
| 52 |
-
"
|
| 53 |
-
"step": 40
|
| 54 |
},
|
| 55 |
{
|
| 56 |
-
"epoch": 0.
|
| 57 |
-
"grad_norm": 0.
|
| 58 |
"learning_rate": 0.0002,
|
| 59 |
-
"loss": 1.
|
| 60 |
-
"step":
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"epoch": 0.
|
| 64 |
-
"grad_norm": 0.
|
| 65 |
"learning_rate": 0.0002,
|
| 66 |
-
"loss": 1.
|
| 67 |
-
"step":
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"epoch": 0.
|
| 71 |
-
"
|
| 72 |
-
"
|
| 73 |
-
"
|
| 74 |
-
"
|
| 75 |
-
"step": 60
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
-
"grad_norm": 0.
|
| 80 |
"learning_rate": 0.0002,
|
| 81 |
-
"loss": 1.
|
| 82 |
-
"step":
|
| 83 |
},
|
| 84 |
{
|
| 85 |
-
"epoch": 0
|
| 86 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
"learning_rate": 0.0002,
|
| 88 |
-
"loss": 1.
|
| 89 |
-
"step":
|
| 90 |
},
|
| 91 |
{
|
| 92 |
-
"epoch":
|
| 93 |
-
"
|
| 94 |
-
"
|
| 95 |
-
"
|
| 96 |
-
"
|
| 97 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 98 |
},
|
| 99 |
{
|
| 100 |
-
"epoch":
|
| 101 |
-
"grad_norm": 0.
|
| 102 |
"learning_rate": 0.0002,
|
| 103 |
-
"loss":
|
| 104 |
-
"step":
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch":
|
| 108 |
-
"grad_norm": 0.
|
| 109 |
"learning_rate": 0.0002,
|
| 110 |
-
"loss":
|
| 111 |
-
"step":
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"
|
| 116 |
-
"
|
| 117 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 118 |
"eval_steps_per_second": 0.599,
|
| 119 |
-
"step":
|
| 120 |
}
|
| 121 |
],
|
| 122 |
-
"logging_steps":
|
| 123 |
-
"max_steps":
|
| 124 |
"num_input_tokens_seen": 0,
|
| 125 |
-
"num_train_epochs":
|
| 126 |
"save_steps": 500,
|
| 127 |
"stateful_callbacks": {
|
| 128 |
"TrainerControl": {
|
|
@@ -136,7 +352,7 @@
|
|
| 136 |
"attributes": {}
|
| 137 |
}
|
| 138 |
},
|
| 139 |
-
"total_flos":
|
| 140 |
"train_batch_size": 4,
|
| 141 |
"trial_name": null,
|
| 142 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 8500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.09411764705882353,
|
| 13 |
+
"grad_norm": 0.30087053775787354,
|
| 14 |
"learning_rate": 0.0002,
|
| 15 |
+
"loss": 1.4482,
|
| 16 |
+
"step": 200
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.18823529411764706,
|
| 20 |
+
"grad_norm": 0.26450973749160767,
|
| 21 |
"learning_rate": 0.0002,
|
| 22 |
+
"loss": 1.3122,
|
| 23 |
+
"step": 400
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.2823529411764706,
|
| 27 |
+
"grad_norm": 0.2747337520122528,
|
| 28 |
+
"learning_rate": 0.0002,
|
| 29 |
+
"loss": 1.2564,
|
| 30 |
+
"step": 600
|
|
|
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.3764705882352941,
|
| 34 |
+
"grad_norm": 0.291003555059433,
|
| 35 |
"learning_rate": 0.0002,
|
| 36 |
+
"loss": 1.2267,
|
| 37 |
+
"step": 800
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.47058823529411764,
|
| 41 |
+
"grad_norm": 0.29030027985572815,
|
| 42 |
"learning_rate": 0.0002,
|
| 43 |
+
"loss": 1.2034,
|
| 44 |
+
"step": 1000
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.5647058823529412,
|
| 48 |
+
"grad_norm": 0.41770946979522705,
|
| 49 |
+
"learning_rate": 0.0002,
|
| 50 |
+
"loss": 1.1856,
|
| 51 |
+
"step": 1200
|
|
|
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.6588235294117647,
|
| 55 |
+
"grad_norm": 0.29782745242118835,
|
| 56 |
"learning_rate": 0.0002,
|
| 57 |
+
"loss": 1.1751,
|
| 58 |
+
"step": 1400
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.7529411764705882,
|
| 62 |
+
"grad_norm": 0.28214672207832336,
|
| 63 |
"learning_rate": 0.0002,
|
| 64 |
+
"loss": 1.1574,
|
| 65 |
+
"step": 1600
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.8470588235294118,
|
| 69 |
+
"grad_norm": 0.29813048243522644,
|
| 70 |
+
"learning_rate": 0.0002,
|
| 71 |
+
"loss": 1.1432,
|
| 72 |
+
"step": 1800
|
|
|
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.9411764705882353,
|
| 76 |
+
"grad_norm": 0.3031373918056488,
|
| 77 |
"learning_rate": 0.0002,
|
| 78 |
+
"loss": 1.1341,
|
| 79 |
+
"step": 2000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 1.0,
|
| 83 |
+
"eval_loss": 1.1254174709320068,
|
| 84 |
+
"eval_runtime": 443.4321,
|
| 85 |
+
"eval_samples_per_second": 4.792,
|
| 86 |
+
"eval_steps_per_second": 0.6,
|
| 87 |
+
"step": 2125
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.035294117647059,
|
| 91 |
+
"grad_norm": 0.30628809332847595,
|
| 92 |
"learning_rate": 0.0002,
|
| 93 |
+
"loss": 1.111,
|
| 94 |
+
"step": 2200
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 1.1294117647058823,
|
| 98 |
+
"grad_norm": 0.3264883756637573,
|
| 99 |
+
"learning_rate": 0.0002,
|
| 100 |
+
"loss": 1.0876,
|
| 101 |
+
"step": 2400
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.223529411764706,
|
| 105 |
+
"grad_norm": 0.3304358720779419,
|
| 106 |
+
"learning_rate": 0.0002,
|
| 107 |
+
"loss": 1.0777,
|
| 108 |
+
"step": 2600
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.3176470588235294,
|
| 112 |
+
"grad_norm": 0.3507118821144104,
|
| 113 |
+
"learning_rate": 0.0002,
|
| 114 |
+
"loss": 1.075,
|
| 115 |
+
"step": 2800
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.4117647058823528,
|
| 119 |
+
"grad_norm": 0.34798240661621094,
|
| 120 |
+
"learning_rate": 0.0002,
|
| 121 |
+
"loss": 1.0705,
|
| 122 |
+
"step": 3000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.5058823529411764,
|
| 126 |
+
"grad_norm": 0.33348146080970764,
|
| 127 |
+
"learning_rate": 0.0002,
|
| 128 |
+
"loss": 1.0616,
|
| 129 |
+
"step": 3200
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.6,
|
| 133 |
+
"grad_norm": 0.3142307698726654,
|
| 134 |
+
"learning_rate": 0.0002,
|
| 135 |
+
"loss": 1.0592,
|
| 136 |
+
"step": 3400
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.6941176470588235,
|
| 140 |
+
"grad_norm": 0.33189332485198975,
|
| 141 |
+
"learning_rate": 0.0002,
|
| 142 |
+
"loss": 1.0563,
|
| 143 |
+
"step": 3600
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.788235294117647,
|
| 147 |
+
"grad_norm": 0.31737592816352844,
|
| 148 |
+
"learning_rate": 0.0002,
|
| 149 |
+
"loss": 1.0508,
|
| 150 |
+
"step": 3800
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.8823529411764706,
|
| 154 |
+
"grad_norm": 0.2998281717300415,
|
| 155 |
+
"learning_rate": 0.0002,
|
| 156 |
+
"loss": 1.0468,
|
| 157 |
+
"step": 4000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.9764705882352942,
|
| 161 |
+
"grad_norm": 0.34619805216789246,
|
| 162 |
+
"learning_rate": 0.0002,
|
| 163 |
+
"loss": 1.0422,
|
| 164 |
+
"step": 4200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 2.0,
|
| 168 |
+
"eval_loss": 1.061407208442688,
|
| 169 |
+
"eval_runtime": 443.4749,
|
| 170 |
+
"eval_samples_per_second": 4.792,
|
| 171 |
+
"eval_steps_per_second": 0.6,
|
| 172 |
+
"step": 4250
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.070588235294118,
|
| 176 |
+
"grad_norm": 0.35598576068878174,
|
| 177 |
+
"learning_rate": 0.0002,
|
| 178 |
+
"loss": 1.0008,
|
| 179 |
+
"step": 4400
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.164705882352941,
|
| 183 |
+
"grad_norm": 0.3873290419578552,
|
| 184 |
+
"learning_rate": 0.0002,
|
| 185 |
+
"loss": 0.9917,
|
| 186 |
+
"step": 4600
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.2588235294117647,
|
| 190 |
+
"grad_norm": 0.3637497127056122,
|
| 191 |
+
"learning_rate": 0.0002,
|
| 192 |
+
"loss": 0.9911,
|
| 193 |
+
"step": 4800
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 2.3529411764705883,
|
| 197 |
+
"grad_norm": 0.35753560066223145,
|
| 198 |
+
"learning_rate": 0.0002,
|
| 199 |
+
"loss": 0.9929,
|
| 200 |
+
"step": 5000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.447058823529412,
|
| 204 |
+
"grad_norm": 0.3278402090072632,
|
| 205 |
+
"learning_rate": 0.0002,
|
| 206 |
+
"loss": 0.9876,
|
| 207 |
+
"step": 5200
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 2.541176470588235,
|
| 211 |
+
"grad_norm": 0.3679386377334595,
|
| 212 |
+
"learning_rate": 0.0002,
|
| 213 |
+
"loss": 0.9842,
|
| 214 |
+
"step": 5400
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 2.635294117647059,
|
| 218 |
+
"grad_norm": 0.3931664526462555,
|
| 219 |
+
"learning_rate": 0.0002,
|
| 220 |
+
"loss": 0.9885,
|
| 221 |
+
"step": 5600
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 2.7294117647058824,
|
| 225 |
+
"grad_norm": 0.3553083539009094,
|
| 226 |
+
"learning_rate": 0.0002,
|
| 227 |
+
"loss": 0.9806,
|
| 228 |
+
"step": 5800
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.8235294117647056,
|
| 232 |
+
"grad_norm": 0.37587428092956543,
|
| 233 |
+
"learning_rate": 0.0002,
|
| 234 |
+
"loss": 0.9796,
|
| 235 |
+
"step": 6000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.9176470588235293,
|
| 239 |
+
"grad_norm": 0.3934173285961151,
|
| 240 |
+
"learning_rate": 0.0002,
|
| 241 |
+
"loss": 0.9786,
|
| 242 |
+
"step": 6200
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 3.0,
|
| 246 |
+
"eval_loss": 1.0245047807693481,
|
| 247 |
+
"eval_runtime": 443.1629,
|
| 248 |
+
"eval_samples_per_second": 4.795,
|
| 249 |
+
"eval_steps_per_second": 0.6,
|
| 250 |
+
"step": 6375
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 3.011764705882353,
|
| 254 |
+
"grad_norm": 0.4304977059364319,
|
| 255 |
+
"learning_rate": 0.0002,
|
| 256 |
+
"loss": 0.9719,
|
| 257 |
+
"step": 6400
|
| 258 |
},
|
| 259 |
{
|
| 260 |
+
"epoch": 3.1058823529411765,
|
| 261 |
+
"grad_norm": 0.39775729179382324,
|
| 262 |
"learning_rate": 0.0002,
|
| 263 |
+
"loss": 0.9174,
|
| 264 |
+
"step": 6600
|
| 265 |
},
|
| 266 |
{
|
| 267 |
+
"epoch": 3.2,
|
| 268 |
+
"grad_norm": 0.40233707427978516,
|
| 269 |
"learning_rate": 0.0002,
|
| 270 |
+
"loss": 0.9271,
|
| 271 |
+
"step": 6800
|
| 272 |
},
|
| 273 |
{
|
| 274 |
+
"epoch": 3.2941176470588234,
|
| 275 |
+
"grad_norm": 0.39777445793151855,
|
| 276 |
+
"learning_rate": 0.0002,
|
| 277 |
+
"loss": 0.9155,
|
| 278 |
+
"step": 7000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 3.388235294117647,
|
| 282 |
+
"grad_norm": 0.4547841548919678,
|
| 283 |
+
"learning_rate": 0.0002,
|
| 284 |
+
"loss": 0.9265,
|
| 285 |
+
"step": 7200
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 3.4823529411764707,
|
| 289 |
+
"grad_norm": 0.3900696039199829,
|
| 290 |
+
"learning_rate": 0.0002,
|
| 291 |
+
"loss": 0.9314,
|
| 292 |
+
"step": 7400
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 3.576470588235294,
|
| 296 |
+
"grad_norm": 0.5135142207145691,
|
| 297 |
+
"learning_rate": 0.0002,
|
| 298 |
+
"loss": 0.9302,
|
| 299 |
+
"step": 7600
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 3.6705882352941175,
|
| 303 |
+
"grad_norm": 0.40233081579208374,
|
| 304 |
+
"learning_rate": 0.0002,
|
| 305 |
+
"loss": 0.9227,
|
| 306 |
+
"step": 7800
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 3.764705882352941,
|
| 310 |
+
"grad_norm": 0.40172523260116577,
|
| 311 |
+
"learning_rate": 0.0002,
|
| 312 |
+
"loss": 0.9268,
|
| 313 |
+
"step": 8000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 3.8588235294117648,
|
| 317 |
+
"grad_norm": 0.38751304149627686,
|
| 318 |
+
"learning_rate": 0.0002,
|
| 319 |
+
"loss": 0.9243,
|
| 320 |
+
"step": 8200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 3.9529411764705884,
|
| 324 |
+
"grad_norm": 0.39530816674232483,
|
| 325 |
+
"learning_rate": 0.0002,
|
| 326 |
+
"loss": 0.9293,
|
| 327 |
+
"step": 8400
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 4.0,
|
| 331 |
+
"eval_loss": 1.0030452013015747,
|
| 332 |
+
"eval_runtime": 443.7686,
|
| 333 |
+
"eval_samples_per_second": 4.789,
|
| 334 |
"eval_steps_per_second": 0.599,
|
| 335 |
+
"step": 8500
|
| 336 |
}
|
| 337 |
],
|
| 338 |
+
"logging_steps": 200,
|
| 339 |
+
"max_steps": 8500,
|
| 340 |
"num_input_tokens_seen": 0,
|
| 341 |
+
"num_train_epochs": 4,
|
| 342 |
"save_steps": 500,
|
| 343 |
"stateful_callbacks": {
|
| 344 |
"TrainerControl": {
|
|
|
|
| 352 |
"attributes": {}
|
| 353 |
}
|
| 354 |
},
|
| 355 |
+
"total_flos": 2.981816312902189e+18,
|
| 356 |
"train_batch_size": 4,
|
| 357 |
"trial_name": null,
|
| 358 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef090e649e75bdbc5f76322c821cf4080d1354b66700ada68766b6363474c05a
|
| 3 |
size 5368
|