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.1751361566.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": 7,
|
| 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:abf304b53a3cc3df48064a1444ff9a1e707542f9d4bf93d212b11cfd2e1381b5
|
| 3 |
+
size 891558696
|
checkpoint-20592/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9415d7a5380b2def8347729b214011bc6585b69a0485934ccf2d996325783fc3
|
| 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:d8c3b78532f81624ed3e9b8de9c5eb5bcf3c25483064d016ca1e22d04b13acba
|
| 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": 0.1858479265338532,
|
| 3 |
+
"best_model_checkpoint": "./logs/mean_codet5-base_L7_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": 3.374912738800049,
|
| 14 |
+
"learning_rate": 7.77302759424796e-06,
|
| 15 |
+
"loss": 2.486,
|
| 16 |
+
"step": 1000
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.777000777000777,
|
| 20 |
+
"grad_norm": 4.941863536834717,
|
| 21 |
+
"learning_rate": 1.554605518849592e-05,
|
| 22 |
+
"loss": 0.5662,
|
| 23 |
+
"step": 2000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.0,
|
| 27 |
+
"eval_bleu": 0.17963100772439886,
|
| 28 |
+
"eval_brevity_penalty": 0.20147036586898603,
|
| 29 |
+
"eval_length_ratio": 0.3843030680069415,
|
| 30 |
+
"eval_loss": 0.3083465099334717,
|
| 31 |
+
"eval_precision": [
|
| 32 |
+
0.9390602368578002,
|
| 33 |
+
0.8998040862020711,
|
| 34 |
+
0.87449401523395,
|
| 35 |
+
0.8552287078392115
|
| 36 |
+
],
|
| 37 |
+
"eval_reference_length": 122164,
|
| 38 |
+
"eval_rouge1": 0.5617603611285253,
|
| 39 |
+
"eval_rouge2": 0.4800176508329015,
|
| 40 |
+
"eval_rougeL": 0.5601128782974164,
|
| 41 |
+
"eval_rougeLsum": 0.5601263101599803,
|
| 42 |
+
"eval_runtime": 18.8064,
|
| 43 |
+
"eval_samples_per_second": 26.533,
|
| 44 |
+
"eval_steps_per_second": 1.702,
|
| 45 |
+
"eval_translation_length": 46948,
|
| 46 |
+
"step": 2574
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.1655011655011656,
|
| 50 |
+
"grad_norm": 3.167290210723877,
|
| 51 |
+
"learning_rate": 1.9631372210471794e-05,
|
| 52 |
+
"loss": 0.3875,
|
| 53 |
+
"step": 3000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.554001554001554,
|
| 57 |
+
"grad_norm": 1.8493537902832031,
|
| 58 |
+
"learning_rate": 1.8768075279492384e-05,
|
| 59 |
+
"loss": 0.3056,
|
| 60 |
+
"step": 4000
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.9425019425019425,
|
| 64 |
+
"grad_norm": 2.211087942123413,
|
| 65 |
+
"learning_rate": 1.790477834851297e-05,
|
| 66 |
+
"loss": 0.2837,
|
| 67 |
+
"step": 5000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 2.0,
|
| 71 |
+
"eval_bleu": 0.1837829816407722,
|
| 72 |
+
"eval_brevity_penalty": 0.20083410383508513,
|
| 73 |
+
"eval_length_ratio": 0.3838364821060214,
|
| 74 |
+
"eval_loss": 0.2173839658498764,
|
| 75 |
+
"eval_precision": [
|
| 76 |
+
0.9480284063039816,
|
| 77 |
+
0.9201802034833592,
|
| 78 |
+
0.9027520536901052,
|
| 79 |
+
0.8904480768383487
|
| 80 |
+
],
|
| 81 |
+
"eval_reference_length": 122164,
|
| 82 |
+
"eval_rouge1": 0.5850236855035034,
|
| 83 |
+
"eval_rouge2": 0.5189301968729537,
|
| 84 |
+
"eval_rougeL": 0.5833269359021196,
|
| 85 |
+
"eval_rougeLsum": 0.5835093022540463,
|
| 86 |
+
"eval_runtime": 18.7121,
|
| 87 |
+
"eval_samples_per_second": 26.667,
|
| 88 |
+
"eval_steps_per_second": 1.71,
|
| 89 |
+
"eval_translation_length": 46891,
|
| 90 |
+
"step": 5148
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 2.331002331002331,
|
| 94 |
+
"grad_norm": 3.8107800483703613,
|
| 95 |
+
"learning_rate": 1.7041481417533562e-05,
|
| 96 |
+
"loss": 0.2162,
|
| 97 |
+
"step": 6000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 2.7195027195027195,
|
| 101 |
+
"grad_norm": 2.2155377864837646,
|
| 102 |
+
"learning_rate": 1.6178184486554153e-05,
|
| 103 |
+
"loss": 0.2066,
|
| 104 |
+
"step": 7000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 3.0,
|
| 108 |
+
"eval_bleu": 0.1854908608571736,
|
| 109 |
+
"eval_brevity_penalty": 0.2013363772102569,
|
| 110 |
+
"eval_length_ratio": 0.38420483939622146,
|
| 111 |
+
"eval_loss": 0.18602398037910461,
|
| 112 |
+
"eval_precision": [
|
| 113 |
+
0.9472473154934379,
|
| 114 |
+
0.9250382238301355,
|
| 115 |
+
0.9116417780486743,
|
| 116 |
+
0.9018904465327142
|
| 117 |
+
],
|
| 118 |
+
"eval_reference_length": 122164,
|
| 119 |
+
"eval_rouge1": 0.5917828855122217,
|
| 120 |
+
"eval_rouge2": 0.5324061960306764,
|
| 121 |
+
"eval_rougeL": 0.5889188371735312,
|
| 122 |
+
"eval_rougeLsum": 0.5891852183701078,
|
| 123 |
+
"eval_runtime": 18.7931,
|
| 124 |
+
"eval_samples_per_second": 26.552,
|
| 125 |
+
"eval_steps_per_second": 1.703,
|
| 126 |
+
"eval_translation_length": 46936,
|
| 127 |
+
"step": 7722
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 3.108003108003108,
|
| 131 |
+
"grad_norm": 0.3534225821495056,
|
| 132 |
+
"learning_rate": 1.5314887555574744e-05,
|
| 133 |
+
"loss": 0.1872,
|
| 134 |
+
"step": 8000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.4965034965034967,
|
| 138 |
+
"grad_norm": 1.0951498746871948,
|
| 139 |
+
"learning_rate": 1.4451590624595331e-05,
|
| 140 |
+
"loss": 0.1505,
|
| 141 |
+
"step": 9000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 3.885003885003885,
|
| 145 |
+
"grad_norm": 1.5113747119903564,
|
| 146 |
+
"learning_rate": 1.358829369361592e-05,
|
| 147 |
+
"loss": 0.1532,
|
| 148 |
+
"step": 10000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 4.0,
|
| 152 |
+
"eval_bleu": 0.18504003653917492,
|
| 153 |
+
"eval_brevity_penalty": 0.1990840203497395,
|
| 154 |
+
"eval_length_ratio": 0.3825513244491012,
|
| 155 |
+
"eval_loss": 0.1719246804714203,
|
| 156 |
+
"eval_precision": [
|
| 157 |
+
0.9552788119998288,
|
| 158 |
+
0.9335568292419163,
|
| 159 |
+
0.9195382193458107,
|
| 160 |
+
0.9100736123085085
|
| 161 |
+
],
|
| 162 |
+
"eval_reference_length": 122164,
|
| 163 |
+
"eval_rouge1": 0.600278595348507,
|
| 164 |
+
"eval_rouge2": 0.5412070759237662,
|
| 165 |
+
"eval_rougeL": 0.5984139762462963,
|
| 166 |
+
"eval_rougeLsum": 0.5974860622159403,
|
| 167 |
+
"eval_runtime": 18.8228,
|
| 168 |
+
"eval_samples_per_second": 26.51,
|
| 169 |
+
"eval_steps_per_second": 1.7,
|
| 170 |
+
"eval_translation_length": 46734,
|
| 171 |
+
"step": 10296
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 4.273504273504273,
|
| 175 |
+
"grad_norm": 0.8724096417427063,
|
| 176 |
+
"learning_rate": 1.272499676263651e-05,
|
| 177 |
+
"loss": 0.1324,
|
| 178 |
+
"step": 11000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 4.662004662004662,
|
| 182 |
+
"grad_norm": 1.0585788488388062,
|
| 183 |
+
"learning_rate": 1.1861699831657098e-05,
|
| 184 |
+
"loss": 0.1242,
|
| 185 |
+
"step": 12000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 5.0,
|
| 189 |
+
"eval_bleu": 0.18522433354346324,
|
| 190 |
+
"eval_brevity_penalty": 0.19899494020906164,
|
| 191 |
+
"eval_length_ratio": 0.3824858387086212,
|
| 192 |
+
"eval_loss": 0.1678268015384674,
|
| 193 |
+
"eval_precision": [
|
| 194 |
+
0.9543508967170312,
|
| 195 |
+
0.9339779782378264,
|
| 196 |
+
0.9217984604618614,
|
| 197 |
+
0.9135731499701519
|
| 198 |
+
],
|
| 199 |
+
"eval_reference_length": 122164,
|
| 200 |
+
"eval_rouge1": 0.5995569817651207,
|
| 201 |
+
"eval_rouge2": 0.5427161978876259,
|
| 202 |
+
"eval_rougeL": 0.5975630584215323,
|
| 203 |
+
"eval_rougeLsum": 0.5977917691117238,
|
| 204 |
+
"eval_runtime": 18.7668,
|
| 205 |
+
"eval_samples_per_second": 26.589,
|
| 206 |
+
"eval_steps_per_second": 1.705,
|
| 207 |
+
"eval_translation_length": 46726,
|
| 208 |
+
"step": 12870
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 5.05050505050505,
|
| 212 |
+
"grad_norm": 1.4060626029968262,
|
| 213 |
+
"learning_rate": 1.0998402900677689e-05,
|
| 214 |
+
"loss": 0.1197,
|
| 215 |
+
"step": 13000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 5.439005439005439,
|
| 219 |
+
"grad_norm": 0.6889358758926392,
|
| 220 |
+
"learning_rate": 1.0135105969698278e-05,
|
| 221 |
+
"loss": 0.1019,
|
| 222 |
+
"step": 14000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 5.827505827505828,
|
| 226 |
+
"grad_norm": 0.4458516538143158,
|
| 227 |
+
"learning_rate": 9.271809038718868e-06,
|
| 228 |
+
"loss": 0.1037,
|
| 229 |
+
"step": 15000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 6.0,
|
| 233 |
+
"eval_bleu": 0.1833010254550757,
|
| 234 |
+
"eval_brevity_penalty": 0.19589418538992515,
|
| 235 |
+
"eval_length_ratio": 0.3802020235093808,
|
| 236 |
+
"eval_loss": 0.16381287574768066,
|
| 237 |
+
"eval_precision": [
|
| 238 |
+
0.9580166641548431,
|
| 239 |
+
0.939018020370854,
|
| 240 |
+
0.9272811282976523,
|
| 241 |
+
0.9189988876529477
|
| 242 |
+
],
|
| 243 |
+
"eval_reference_length": 122164,
|
| 244 |
+
"eval_rouge1": 0.6040698008982983,
|
| 245 |
+
"eval_rouge2": 0.5499471553858646,
|
| 246 |
+
"eval_rougeL": 0.6023900379269751,
|
| 247 |
+
"eval_rougeLsum": 0.6023409541861235,
|
| 248 |
+
"eval_runtime": 18.7493,
|
| 249 |
+
"eval_samples_per_second": 26.614,
|
| 250 |
+
"eval_steps_per_second": 1.707,
|
| 251 |
+
"eval_translation_length": 46447,
|
| 252 |
+
"step": 15444
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 6.216006216006216,
|
| 256 |
+
"grad_norm": 0.46098294854164124,
|
| 257 |
+
"learning_rate": 8.408512107739457e-06,
|
| 258 |
+
"loss": 0.0915,
|
| 259 |
+
"step": 16000
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 6.6045066045066045,
|
| 263 |
+
"grad_norm": 1.5093261003494263,
|
| 264 |
+
"learning_rate": 7.545215176760047e-06,
|
| 265 |
+
"loss": 0.0872,
|
| 266 |
+
"step": 17000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.993006993006993,
|
| 270 |
+
"grad_norm": 0.5046265721321106,
|
| 271 |
+
"learning_rate": 6.681918245780636e-06,
|
| 272 |
+
"loss": 0.0861,
|
| 273 |
+
"step": 18000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 7.0,
|
| 277 |
+
"eval_bleu": 0.1829645975661277,
|
| 278 |
+
"eval_brevity_penalty": 0.1944201518358197,
|
| 279 |
+
"eval_length_ratio": 0.37911332307390067,
|
| 280 |
+
"eval_loss": 0.16533204913139343,
|
| 281 |
+
"eval_precision": [
|
| 282 |
+
0.9612644124886643,
|
| 283 |
+
0.9441667576121358,
|
| 284 |
+
0.9333127372230559,
|
| 285 |
+
0.9259432804516143
|
| 286 |
+
],
|
| 287 |
+
"eval_reference_length": 122164,
|
| 288 |
+
"eval_rouge1": 0.606793175048812,
|
| 289 |
+
"eval_rouge2": 0.5538059035055065,
|
| 290 |
+
"eval_rougeL": 0.6049968026886056,
|
| 291 |
+
"eval_rougeLsum": 0.6048344970026875,
|
| 292 |
+
"eval_runtime": 18.7907,
|
| 293 |
+
"eval_samples_per_second": 26.556,
|
| 294 |
+
"eval_steps_per_second": 1.703,
|
| 295 |
+
"eval_translation_length": 46314,
|
| 296 |
+
"step": 18018
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.381507381507381,
|
| 300 |
+
"grad_norm": 1.0257986783981323,
|
| 301 |
+
"learning_rate": 5.818621314801226e-06,
|
| 302 |
+
"loss": 0.0771,
|
| 303 |
+
"step": 19000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 7.77000777000777,
|
| 307 |
+
"grad_norm": 1.0541136264801025,
|
| 308 |
+
"learning_rate": 4.955324383821816e-06,
|
| 309 |
+
"loss": 0.0696,
|
| 310 |
+
"step": 20000
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 8.0,
|
| 314 |
+
"eval_bleu": 0.1858479265338532,
|
| 315 |
+
"eval_brevity_penalty": 0.1994961376851703,
|
| 316 |
+
"eval_length_ratio": 0.38285419599882126,
|
| 317 |
+
"eval_loss": 0.16029638051986694,
|
| 318 |
+
"eval_precision": [
|
| 319 |
+
0.9521498364371085,
|
| 320 |
+
0.9344528008298755,
|
| 321 |
+
0.9237323312870033,
|
| 322 |
+
0.9163979325882405
|
| 323 |
+
],
|
| 324 |
+
"eval_reference_length": 122164,
|
| 325 |
+
"eval_rouge1": 0.599509561418239,
|
| 326 |
+
"eval_rouge2": 0.5462969370634634,
|
| 327 |
+
"eval_rougeL": 0.5977257138698633,
|
| 328 |
+
"eval_rougeLsum": 0.59798598702676,
|
| 329 |
+
"eval_runtime": 18.7364,
|
| 330 |
+
"eval_samples_per_second": 26.633,
|
| 331 |
+
"eval_steps_per_second": 1.708,
|
| 332 |
+
"eval_translation_length": 46771,
|
| 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 sha256:415ec8809d2902c7e877f6a642a826fc09e949dca1db00db7b019628c3cd92a7
|
| 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": 7,
|
| 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.1751361566.gra5
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:148ef7eea982c154ba189ff85ac1c723389f668b2e606217a00e795e6d10716e
|
| 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:abf304b53a3cc3df48064a1444ff9a1e707542f9d4bf93d212b11cfd2e1381b5
|
| 3 |
+
size 891558696
|
results.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"rouge1": 0.7403387813276292,
|
| 3 |
+
"rouge2": 0.692668617987129,
|
| 4 |
+
"rougeL": 0.7380851221622119,
|
| 5 |
+
"rougeLsum": 0.7383027414500516,
|
| 6 |
+
"bleu": 0.6601602587714013,
|
| 7 |
+
"precision": [
|
| 8 |
+
0.6802433588631092,
|
| 9 |
+
0.6625922164582171,
|
| 10 |
+
0.6528111301113194,
|
| 11 |
+
0.6455045612570843
|
| 12 |
+
],
|
| 13 |
+
"brevity_penalty": 1.0,
|
| 14 |
+
"length_ratio": 1.4444629877687256,
|
| 15 |
+
"translation_length": 328897,
|
| 16 |
+
"reference_length": 227695
|
| 17 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:415ec8809d2902c7e877f6a642a826fc09e949dca1db00db7b019628c3cd92a7
|
| 3 |
+
size 5560
|