Upload folder using huggingface_hub
Browse files- checkpoint-20592/config.json +71 -0
- checkpoint-20592/generation_config.json +7 -0
- checkpoint-20592/model.safetensors +3 -0
- checkpoint-20592/optimizer.pt +3 -0
- checkpoint-20592/rng_state.pth +3 -0
- checkpoint-20592/scheduler.pt +3 -0
- checkpoint-20592/trainer_state.json +357 -0
- checkpoint-20592/training_args.bin +3 -0
- config.json +71 -0
- events.out.tfevents.1751355165.gra5 +3 -0
- generation_config.json +7 -0
- model.safetensors +3 -0
- results.json +17 -0
- training_args.bin +3 -0
checkpoint-20592/config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/home/mootez/projects/def-tusharma/mootez/local_models/codet5-base",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"T5ForConditionalGeneration"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": 0.0,
|
| 8 |
+
"d_ff": 3072,
|
| 9 |
+
"d_kv": 64,
|
| 10 |
+
"d_model": 768,
|
| 11 |
+
"decoder_start_token_id": 0,
|
| 12 |
+
"dense_act_fn": "relu",
|
| 13 |
+
"dropout_rate": 0.1,
|
| 14 |
+
"eos_token_id": 2,
|
| 15 |
+
"feed_forward_proj": "relu",
|
| 16 |
+
"gradient_checkpointing": false,
|
| 17 |
+
"id2label": {
|
| 18 |
+
"0": "LABEL_0"
|
| 19 |
+
},
|
| 20 |
+
"initializer_factor": 1.0,
|
| 21 |
+
"is_encoder_decoder": true,
|
| 22 |
+
"is_gated_act": false,
|
| 23 |
+
"label2id": {
|
| 24 |
+
"LABEL_0": 0
|
| 25 |
+
},
|
| 26 |
+
"layer_norm_epsilon": 1e-06,
|
| 27 |
+
"merge_layer_index": 6,
|
| 28 |
+
"merger_type": "mean",
|
| 29 |
+
"model_type": "t5",
|
| 30 |
+
"n_positions": 512,
|
| 31 |
+
"num_decoder_layers": 12,
|
| 32 |
+
"num_heads": 12,
|
| 33 |
+
"num_layers": 12,
|
| 34 |
+
"output_past": true,
|
| 35 |
+
"pad_token_id": 0,
|
| 36 |
+
"relative_attention_max_distance": 128,
|
| 37 |
+
"relative_attention_num_buckets": 32,
|
| 38 |
+
"task_specific_params": {
|
| 39 |
+
"summarization": {
|
| 40 |
+
"early_stopping": true,
|
| 41 |
+
"length_penalty": 2.0,
|
| 42 |
+
"max_length": 200,
|
| 43 |
+
"min_length": 30,
|
| 44 |
+
"no_repeat_ngram_size": 3,
|
| 45 |
+
"num_beams": 4,
|
| 46 |
+
"prefix": "summarize: "
|
| 47 |
+
},
|
| 48 |
+
"translation_en_to_de": {
|
| 49 |
+
"early_stopping": true,
|
| 50 |
+
"max_length": 300,
|
| 51 |
+
"num_beams": 4,
|
| 52 |
+
"prefix": "translate English to German: "
|
| 53 |
+
},
|
| 54 |
+
"translation_en_to_fr": {
|
| 55 |
+
"early_stopping": true,
|
| 56 |
+
"max_length": 300,
|
| 57 |
+
"num_beams": 4,
|
| 58 |
+
"prefix": "translate English to French: "
|
| 59 |
+
},
|
| 60 |
+
"translation_en_to_ro": {
|
| 61 |
+
"early_stopping": true,
|
| 62 |
+
"max_length": 300,
|
| 63 |
+
"num_beams": 4,
|
| 64 |
+
"prefix": "translate English to Romanian: "
|
| 65 |
+
}
|
| 66 |
+
},
|
| 67 |
+
"torch_dtype": "float32",
|
| 68 |
+
"transformers_version": "4.48.2",
|
| 69 |
+
"use_cache": true,
|
| 70 |
+
"vocab_size": 32100
|
| 71 |
+
}
|
checkpoint-20592/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 1,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.48.2"
|
| 7 |
+
}
|
checkpoint-20592/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4cc8a60aa4c781fc51052ebf97673a148db0faa68f971463ae82e4db0e737d84
|
| 3 |
+
size 891558696
|
checkpoint-20592/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e78f0db78cf3f88ee8b411e79d1beb21c4ebde1d7b52eec0ef8e557bd633dbd2
|
| 3 |
+
size 1783272762
|
checkpoint-20592/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7aa8f5e9aab948dbfcd3d6a71008df888b57cbd3bffa0f801d1ac7deb7f19889
|
| 3 |
+
size 14244
|
checkpoint-20592/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b6652c945adbf45ef008e787a68f45797ae62e10f4915e2e70239750b0bd4aa
|
| 3 |
+
size 1064
|
checkpoint-20592/trainer_state.json
ADDED
|
@@ -0,0 +1,357 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.1837775804231829,
|
| 3 |
+
"best_model_checkpoint": "./logs/mean_codet5-base_L6_codetrans_123456_src_java_tgt_c_sharp/checkpoint-20592",
|
| 4 |
+
"epoch": 8.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 20592,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.3885003885003885,
|
| 13 |
+
"grad_norm": 4.471341609954834,
|
| 14 |
+
"learning_rate": 7.77302759424796e-06,
|
| 15 |
+
"loss": 2.616,
|
| 16 |
+
"step": 1000
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.777000777000777,
|
| 20 |
+
"grad_norm": 4.842982292175293,
|
| 21 |
+
"learning_rate": 1.554605518849592e-05,
|
| 22 |
+
"loss": 0.5908,
|
| 23 |
+
"step": 2000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.0,
|
| 27 |
+
"eval_bleu": 0.17937504468424068,
|
| 28 |
+
"eval_brevity_penalty": 0.20109078498001373,
|
| 29 |
+
"eval_length_ratio": 0.38402475360990146,
|
| 30 |
+
"eval_loss": 0.31831929087638855,
|
| 31 |
+
"eval_precision": [
|
| 32 |
+
0.9407426354606301,
|
| 33 |
+
0.9008725627491113,
|
| 34 |
+
0.8744664169352732,
|
| 35 |
+
0.8542836382852236
|
| 36 |
+
],
|
| 37 |
+
"eval_reference_length": 122164,
|
| 38 |
+
"eval_rouge1": 0.5614755980593795,
|
| 39 |
+
"eval_rouge2": 0.476325016461125,
|
| 40 |
+
"eval_rougeL": 0.5599105046866861,
|
| 41 |
+
"eval_rougeLsum": 0.5597356395582445,
|
| 42 |
+
"eval_runtime": 18.9208,
|
| 43 |
+
"eval_samples_per_second": 26.373,
|
| 44 |
+
"eval_steps_per_second": 1.691,
|
| 45 |
+
"eval_translation_length": 46914,
|
| 46 |
+
"step": 2574
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.1655011655011656,
|
| 50 |
+
"grad_norm": 3.1602447032928467,
|
| 51 |
+
"learning_rate": 1.9631372210471794e-05,
|
| 52 |
+
"loss": 0.4003,
|
| 53 |
+
"step": 3000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.554001554001554,
|
| 57 |
+
"grad_norm": 1.2096540927886963,
|
| 58 |
+
"learning_rate": 1.8768075279492384e-05,
|
| 59 |
+
"loss": 0.3131,
|
| 60 |
+
"step": 4000
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.9425019425019425,
|
| 64 |
+
"grad_norm": 2.7136287689208984,
|
| 65 |
+
"learning_rate": 1.790477834851297e-05,
|
| 66 |
+
"loss": 0.287,
|
| 67 |
+
"step": 5000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 2.0,
|
| 71 |
+
"eval_bleu": 0.18362196066352315,
|
| 72 |
+
"eval_brevity_penalty": 0.2005328797271596,
|
| 73 |
+
"eval_length_ratio": 0.38361546773190136,
|
| 74 |
+
"eval_loss": 0.2202845811843872,
|
| 75 |
+
"eval_precision": [
|
| 76 |
+
0.9497482075793786,
|
| 77 |
+
0.9206729213846652,
|
| 78 |
+
0.9030654515327258,
|
| 79 |
+
0.8902726651530849
|
| 80 |
+
],
|
| 81 |
+
"eval_reference_length": 122164,
|
| 82 |
+
"eval_rouge1": 0.5867295530888541,
|
| 83 |
+
"eval_rouge2": 0.5182796875384879,
|
| 84 |
+
"eval_rougeL": 0.5836223189870129,
|
| 85 |
+
"eval_rougeLsum": 0.5834809642116414,
|
| 86 |
+
"eval_runtime": 18.84,
|
| 87 |
+
"eval_samples_per_second": 26.486,
|
| 88 |
+
"eval_steps_per_second": 1.699,
|
| 89 |
+
"eval_translation_length": 46864,
|
| 90 |
+
"step": 5148
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 2.331002331002331,
|
| 94 |
+
"grad_norm": 4.927639961242676,
|
| 95 |
+
"learning_rate": 1.7041481417533562e-05,
|
| 96 |
+
"loss": 0.2208,
|
| 97 |
+
"step": 6000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 2.7195027195027195,
|
| 101 |
+
"grad_norm": 2.7160425186157227,
|
| 102 |
+
"learning_rate": 1.6178184486554153e-05,
|
| 103 |
+
"loss": 0.211,
|
| 104 |
+
"step": 7000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 3.0,
|
| 108 |
+
"eval_bleu": 0.1816887032840445,
|
| 109 |
+
"eval_brevity_penalty": 0.19657113853740538,
|
| 110 |
+
"eval_length_ratio": 0.3807013522805409,
|
| 111 |
+
"eval_loss": 0.19259938597679138,
|
| 112 |
+
"eval_precision": [
|
| 113 |
+
0.954265932742754,
|
| 114 |
+
0.9281444934686691,
|
| 115 |
+
0.9131619424302351,
|
| 116 |
+
0.9024016351558508
|
| 117 |
+
],
|
| 118 |
+
"eval_reference_length": 122164,
|
| 119 |
+
"eval_rouge1": 0.5900312520080135,
|
| 120 |
+
"eval_rouge2": 0.5257477510867983,
|
| 121 |
+
"eval_rougeL": 0.5878283588531475,
|
| 122 |
+
"eval_rougeLsum": 0.5877529941718853,
|
| 123 |
+
"eval_runtime": 19.0038,
|
| 124 |
+
"eval_samples_per_second": 26.258,
|
| 125 |
+
"eval_steps_per_second": 1.684,
|
| 126 |
+
"eval_translation_length": 46508,
|
| 127 |
+
"step": 7722
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 3.108003108003108,
|
| 131 |
+
"grad_norm": 0.5720973610877991,
|
| 132 |
+
"learning_rate": 1.5314887555574744e-05,
|
| 133 |
+
"loss": 0.1904,
|
| 134 |
+
"step": 8000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.4965034965034967,
|
| 138 |
+
"grad_norm": 1.2556794881820679,
|
| 139 |
+
"learning_rate": 1.4451590624595331e-05,
|
| 140 |
+
"loss": 0.1534,
|
| 141 |
+
"step": 9000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 3.885003885003885,
|
| 145 |
+
"grad_norm": 1.4727575778961182,
|
| 146 |
+
"learning_rate": 1.358829369361592e-05,
|
| 147 |
+
"loss": 0.1547,
|
| 148 |
+
"step": 10000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 4.0,
|
| 152 |
+
"eval_bleu": 0.18279275962070415,
|
| 153 |
+
"eval_brevity_penalty": 0.19589418538992515,
|
| 154 |
+
"eval_length_ratio": 0.3802020235093808,
|
| 155 |
+
"eval_loss": 0.17794129252433777,
|
| 156 |
+
"eval_precision": [
|
| 157 |
+
0.9590931599457446,
|
| 158 |
+
0.9367763558805606,
|
| 159 |
+
0.923342647803032,
|
| 160 |
+
0.9138820912124583
|
| 161 |
+
],
|
| 162 |
+
"eval_reference_length": 122164,
|
| 163 |
+
"eval_rouge1": 0.5975943301954172,
|
| 164 |
+
"eval_rouge2": 0.5388068788629989,
|
| 165 |
+
"eval_rougeL": 0.5965036732643416,
|
| 166 |
+
"eval_rougeLsum": 0.5955866849866471,
|
| 167 |
+
"eval_runtime": 19.0048,
|
| 168 |
+
"eval_samples_per_second": 26.256,
|
| 169 |
+
"eval_steps_per_second": 1.684,
|
| 170 |
+
"eval_translation_length": 46447,
|
| 171 |
+
"step": 10296
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 4.273504273504273,
|
| 175 |
+
"grad_norm": 0.934843897819519,
|
| 176 |
+
"learning_rate": 1.272499676263651e-05,
|
| 177 |
+
"loss": 0.134,
|
| 178 |
+
"step": 11000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 4.662004662004662,
|
| 182 |
+
"grad_norm": 1.171906590461731,
|
| 183 |
+
"learning_rate": 1.1861699831657098e-05,
|
| 184 |
+
"loss": 0.1282,
|
| 185 |
+
"step": 12000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 5.0,
|
| 189 |
+
"eval_bleu": 0.1830765775922422,
|
| 190 |
+
"eval_brevity_penalty": 0.19541732524551025,
|
| 191 |
+
"eval_length_ratio": 0.37985003765430075,
|
| 192 |
+
"eval_loss": 0.16766929626464844,
|
| 193 |
+
"eval_precision": [
|
| 194 |
+
0.9608007930350831,
|
| 195 |
+
0.9405511382202374,
|
| 196 |
+
0.9277188036823327,
|
| 197 |
+
0.9188545215667936
|
| 198 |
+
],
|
| 199 |
+
"eval_reference_length": 122164,
|
| 200 |
+
"eval_rouge1": 0.6018337480835432,
|
| 201 |
+
"eval_rouge2": 0.5446167801077577,
|
| 202 |
+
"eval_rougeL": 0.5999491777726305,
|
| 203 |
+
"eval_rougeLsum": 0.6001372300301927,
|
| 204 |
+
"eval_runtime": 18.9569,
|
| 205 |
+
"eval_samples_per_second": 26.323,
|
| 206 |
+
"eval_steps_per_second": 1.688,
|
| 207 |
+
"eval_translation_length": 46404,
|
| 208 |
+
"step": 12870
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 5.05050505050505,
|
| 212 |
+
"grad_norm": 1.4828258752822876,
|
| 213 |
+
"learning_rate": 1.0998402900677689e-05,
|
| 214 |
+
"loss": 0.1218,
|
| 215 |
+
"step": 13000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 5.439005439005439,
|
| 219 |
+
"grad_norm": 0.31411540508270264,
|
| 220 |
+
"learning_rate": 1.0135105969698278e-05,
|
| 221 |
+
"loss": 0.104,
|
| 222 |
+
"step": 14000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 5.827505827505828,
|
| 226 |
+
"grad_norm": 1.1318721771240234,
|
| 227 |
+
"learning_rate": 9.271809038718868e-06,
|
| 228 |
+
"loss": 0.1054,
|
| 229 |
+
"step": 15000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 6.0,
|
| 233 |
+
"eval_bleu": 0.1835051319254437,
|
| 234 |
+
"eval_brevity_penalty": 0.19553928617025274,
|
| 235 |
+
"eval_length_ratio": 0.3799400805474608,
|
| 236 |
+
"eval_loss": 0.16865986585617065,
|
| 237 |
+
"eval_precision": [
|
| 238 |
+
0.9609824410212215,
|
| 239 |
+
0.9419156720968725,
|
| 240 |
+
0.9299161107074444,
|
| 241 |
+
0.9214791397657954
|
| 242 |
+
],
|
| 243 |
+
"eval_reference_length": 122164,
|
| 244 |
+
"eval_rouge1": 0.6046164832194536,
|
| 245 |
+
"eval_rouge2": 0.5489603519606353,
|
| 246 |
+
"eval_rougeL": 0.6027727929472609,
|
| 247 |
+
"eval_rougeLsum": 0.6026512950008954,
|
| 248 |
+
"eval_runtime": 18.8243,
|
| 249 |
+
"eval_samples_per_second": 26.508,
|
| 250 |
+
"eval_steps_per_second": 1.7,
|
| 251 |
+
"eval_translation_length": 46415,
|
| 252 |
+
"step": 15444
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 6.216006216006216,
|
| 256 |
+
"grad_norm": 1.9530174732208252,
|
| 257 |
+
"learning_rate": 8.408512107739457e-06,
|
| 258 |
+
"loss": 0.0927,
|
| 259 |
+
"step": 16000
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 6.6045066045066045,
|
| 263 |
+
"grad_norm": 1.5964767932891846,
|
| 264 |
+
"learning_rate": 7.545215176760047e-06,
|
| 265 |
+
"loss": 0.088,
|
| 266 |
+
"step": 17000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.993006993006993,
|
| 270 |
+
"grad_norm": 0.626581072807312,
|
| 271 |
+
"learning_rate": 6.681918245780636e-06,
|
| 272 |
+
"loss": 0.0872,
|
| 273 |
+
"step": 18000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 7.0,
|
| 277 |
+
"eval_bleu": 0.18337063512511098,
|
| 278 |
+
"eval_brevity_penalty": 0.19533972326857418,
|
| 279 |
+
"eval_length_ratio": 0.3797927376313808,
|
| 280 |
+
"eval_loss": 0.16658362746238708,
|
| 281 |
+
"eval_precision": [
|
| 282 |
+
0.9602129448024657,
|
| 283 |
+
0.9418493180530743,
|
| 284 |
+
0.9305491310381286,
|
| 285 |
+
0.92271714922049
|
| 286 |
+
],
|
| 287 |
+
"eval_reference_length": 122164,
|
| 288 |
+
"eval_rouge1": 0.6045784085256708,
|
| 289 |
+
"eval_rouge2": 0.549897984490922,
|
| 290 |
+
"eval_rougeL": 0.6029606153603441,
|
| 291 |
+
"eval_rougeLsum": 0.6027976132042671,
|
| 292 |
+
"eval_runtime": 18.8951,
|
| 293 |
+
"eval_samples_per_second": 26.409,
|
| 294 |
+
"eval_steps_per_second": 1.694,
|
| 295 |
+
"eval_translation_length": 46397,
|
| 296 |
+
"step": 18018
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.381507381507381,
|
| 300 |
+
"grad_norm": 1.594728946685791,
|
| 301 |
+
"learning_rate": 5.818621314801226e-06,
|
| 302 |
+
"loss": 0.0788,
|
| 303 |
+
"step": 19000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 7.77000777000777,
|
| 307 |
+
"grad_norm": 1.2195911407470703,
|
| 308 |
+
"learning_rate": 4.955324383821816e-06,
|
| 309 |
+
"loss": 0.0719,
|
| 310 |
+
"step": 20000
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 8.0,
|
| 314 |
+
"eval_bleu": 0.1837775804231829,
|
| 315 |
+
"eval_brevity_penalty": 0.19601621761598836,
|
| 316 |
+
"eval_length_ratio": 0.3802920664025409,
|
| 317 |
+
"eval_loss": 0.16370242834091187,
|
| 318 |
+
"eval_precision": [
|
| 319 |
+
0.9584571010374963,
|
| 320 |
+
0.940338127461433,
|
| 321 |
+
0.9296524417069951,
|
| 322 |
+
0.922199239340762
|
| 323 |
+
],
|
| 324 |
+
"eval_reference_length": 122164,
|
| 325 |
+
"eval_rouge1": 0.604076479558441,
|
| 326 |
+
"eval_rouge2": 0.5498539552524238,
|
| 327 |
+
"eval_rougeL": 0.602226678445738,
|
| 328 |
+
"eval_rougeLsum": 0.6023887033775726,
|
| 329 |
+
"eval_runtime": 18.7993,
|
| 330 |
+
"eval_samples_per_second": 26.543,
|
| 331 |
+
"eval_steps_per_second": 1.702,
|
| 332 |
+
"eval_translation_length": 46458,
|
| 333 |
+
"step": 20592
|
| 334 |
+
}
|
| 335 |
+
],
|
| 336 |
+
"logging_steps": 1000,
|
| 337 |
+
"max_steps": 25740,
|
| 338 |
+
"num_input_tokens_seen": 0,
|
| 339 |
+
"num_train_epochs": 10,
|
| 340 |
+
"save_steps": 500,
|
| 341 |
+
"stateful_callbacks": {
|
| 342 |
+
"TrainerControl": {
|
| 343 |
+
"args": {
|
| 344 |
+
"should_epoch_stop": false,
|
| 345 |
+
"should_evaluate": false,
|
| 346 |
+
"should_log": false,
|
| 347 |
+
"should_save": true,
|
| 348 |
+
"should_training_stop": false
|
| 349 |
+
},
|
| 350 |
+
"attributes": {}
|
| 351 |
+
}
|
| 352 |
+
},
|
| 353 |
+
"total_flos": 2.4978929347584e+16,
|
| 354 |
+
"train_batch_size": 4,
|
| 355 |
+
"trial_name": null,
|
| 356 |
+
"trial_params": null
|
| 357 |
+
}
|
checkpoint-20592/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33816964e8c6fe2a44d96dc850c903495fbaf84433aefb247a6800a62d53382a
|
| 3 |
+
size 5560
|
config.json
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/home/mootez/projects/def-tusharma/mootez/local_models/codet5-base",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"T5ForConditionalGeneration"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"classifier_dropout": 0.0,
|
| 8 |
+
"d_ff": 3072,
|
| 9 |
+
"d_kv": 64,
|
| 10 |
+
"d_model": 768,
|
| 11 |
+
"decoder_start_token_id": 0,
|
| 12 |
+
"dense_act_fn": "relu",
|
| 13 |
+
"dropout_rate": 0.1,
|
| 14 |
+
"eos_token_id": 2,
|
| 15 |
+
"feed_forward_proj": "relu",
|
| 16 |
+
"gradient_checkpointing": false,
|
| 17 |
+
"id2label": {
|
| 18 |
+
"0": "LABEL_0"
|
| 19 |
+
},
|
| 20 |
+
"initializer_factor": 1.0,
|
| 21 |
+
"is_encoder_decoder": true,
|
| 22 |
+
"is_gated_act": false,
|
| 23 |
+
"label2id": {
|
| 24 |
+
"LABEL_0": 0
|
| 25 |
+
},
|
| 26 |
+
"layer_norm_epsilon": 1e-06,
|
| 27 |
+
"merge_layer_index": 6,
|
| 28 |
+
"merger_type": "mean",
|
| 29 |
+
"model_type": "t5",
|
| 30 |
+
"n_positions": 512,
|
| 31 |
+
"num_decoder_layers": 12,
|
| 32 |
+
"num_heads": 12,
|
| 33 |
+
"num_layers": 12,
|
| 34 |
+
"output_past": true,
|
| 35 |
+
"pad_token_id": 0,
|
| 36 |
+
"relative_attention_max_distance": 128,
|
| 37 |
+
"relative_attention_num_buckets": 32,
|
| 38 |
+
"task_specific_params": {
|
| 39 |
+
"summarization": {
|
| 40 |
+
"early_stopping": true,
|
| 41 |
+
"length_penalty": 2.0,
|
| 42 |
+
"max_length": 200,
|
| 43 |
+
"min_length": 30,
|
| 44 |
+
"no_repeat_ngram_size": 3,
|
| 45 |
+
"num_beams": 4,
|
| 46 |
+
"prefix": "summarize: "
|
| 47 |
+
},
|
| 48 |
+
"translation_en_to_de": {
|
| 49 |
+
"early_stopping": true,
|
| 50 |
+
"max_length": 300,
|
| 51 |
+
"num_beams": 4,
|
| 52 |
+
"prefix": "translate English to German: "
|
| 53 |
+
},
|
| 54 |
+
"translation_en_to_fr": {
|
| 55 |
+
"early_stopping": true,
|
| 56 |
+
"max_length": 300,
|
| 57 |
+
"num_beams": 4,
|
| 58 |
+
"prefix": "translate English to French: "
|
| 59 |
+
},
|
| 60 |
+
"translation_en_to_ro": {
|
| 61 |
+
"early_stopping": true,
|
| 62 |
+
"max_length": 300,
|
| 63 |
+
"num_beams": 4,
|
| 64 |
+
"prefix": "translate English to Romanian: "
|
| 65 |
+
}
|
| 66 |
+
},
|
| 67 |
+
"torch_dtype": "float32",
|
| 68 |
+
"transformers_version": "4.48.2",
|
| 69 |
+
"use_cache": true,
|
| 70 |
+
"vocab_size": 32100
|
| 71 |
+
}
|
events.out.tfevents.1751355165.gra5
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c037f14acc32bcc61bc57a1192fb956c169d85ad15819f18d588681cb7dfb3fa
|
| 3 |
+
size 19672
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 1,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.48.2"
|
| 7 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4cc8a60aa4c781fc51052ebf97673a148db0faa68f971463ae82e4db0e737d84
|
| 3 |
+
size 891558696
|
results.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"rouge1": 0.7627668273394363,
|
| 3 |
+
"rouge2": 0.714474944930078,
|
| 4 |
+
"rougeL": 0.7602423163535736,
|
| 5 |
+
"rougeLsum": 0.7601341743814951,
|
| 6 |
+
"bleu": 0.6808604451475841,
|
| 7 |
+
"precision": [
|
| 8 |
+
0.7020455265278731,
|
| 9 |
+
0.6833757482910809,
|
| 10 |
+
0.673083173709572,
|
| 11 |
+
0.6654858796230345
|
| 12 |
+
],
|
| 13 |
+
"brevity_penalty": 1.0,
|
| 14 |
+
"length_ratio": 1.4005182371154394,
|
| 15 |
+
"translation_length": 318891,
|
| 16 |
+
"reference_length": 227695
|
| 17 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33816964e8c6fe2a44d96dc850c903495fbaf84433aefb247a6800a62d53382a
|
| 3 |
+
size 5560
|