Upload folder using huggingface_hub
Browse files- checkpoint-25740/config.json +71 -0
- checkpoint-25740/generation_config.json +7 -0
- checkpoint-25740/model.safetensors +3 -0
- checkpoint-25740/optimizer.pt +3 -0
- checkpoint-25740/rng_state.pth +3 -0
- checkpoint-25740/scheduler.pt +3 -0
- checkpoint-25740/trainer_state.json +438 -0
- checkpoint-25740/training_args.bin +3 -0
- config.json +71 -0
- events.out.tfevents.1751331117.gra5 +3 -0
- generation_config.json +7 -0
- model.safetensors +3 -0
- results.json +17 -0
- training_args.bin +3 -0
checkpoint-25740/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": 0,
|
| 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-25740/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-25740/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3204e3f87f0aa024b6c789452cc8948835758c7e1220ddba095a680a7121e2f
|
| 3 |
+
size 891558696
|
checkpoint-25740/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:44f885ff4a988de10477ef40af01fe940e966fe07859b4fa1b7a629869329254
|
| 3 |
+
size 1783272762
|
checkpoint-25740/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37854a36f2116d7d1c7da1f40f2bea25a89cc1460689e8c6811d7adf1b5c0f5a
|
| 3 |
+
size 14244
|
checkpoint-25740/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81d4813478af8012e8386fe79481b7be6da5268dd2e71b74f5e3cad0d4429beb
|
| 3 |
+
size 1064
|
checkpoint-25740/trainer_state.json
ADDED
|
@@ -0,0 +1,438 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.1853679051126517,
|
| 3 |
+
"best_model_checkpoint": "./logs/mean_codet5-base_L0_codetrans_123456_src_java_tgt_c_sharp/checkpoint-25740",
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 25740,
|
| 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": 5.230306148529053,
|
| 14 |
+
"learning_rate": 7.77302759424796e-06,
|
| 15 |
+
"loss": 3.1643,
|
| 16 |
+
"step": 1000
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.777000777000777,
|
| 20 |
+
"grad_norm": 5.397551536560059,
|
| 21 |
+
"learning_rate": 1.554605518849592e-05,
|
| 22 |
+
"loss": 0.7168,
|
| 23 |
+
"step": 2000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.0,
|
| 27 |
+
"eval_bleu": 0.17734586078415732,
|
| 28 |
+
"eval_brevity_penalty": 0.2019618338664502,
|
| 29 |
+
"eval_length_ratio": 0.38466323957958154,
|
| 30 |
+
"eval_loss": 0.3681968152523041,
|
| 31 |
+
"eval_precision": [
|
| 32 |
+
0.9374361593462717,
|
| 33 |
+
0.8884563267588669,
|
| 34 |
+
0.8572857329216854,
|
| 35 |
+
0.832728871304539
|
| 36 |
+
],
|
| 37 |
+
"eval_reference_length": 122164,
|
| 38 |
+
"eval_rouge1": 0.5427229679888004,
|
| 39 |
+
"eval_rouge2": 0.4446765009777419,
|
| 40 |
+
"eval_rougeL": 0.5403723729065508,
|
| 41 |
+
"eval_rougeLsum": 0.5401834271328774,
|
| 42 |
+
"eval_runtime": 18.984,
|
| 43 |
+
"eval_samples_per_second": 26.285,
|
| 44 |
+
"eval_steps_per_second": 1.686,
|
| 45 |
+
"eval_translation_length": 46992,
|
| 46 |
+
"step": 2574
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 1.1655011655011656,
|
| 50 |
+
"grad_norm": 6.40755033493042,
|
| 51 |
+
"learning_rate": 1.9631372210471794e-05,
|
| 52 |
+
"loss": 0.4834,
|
| 53 |
+
"step": 3000
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.554001554001554,
|
| 57 |
+
"grad_norm": 2.0518579483032227,
|
| 58 |
+
"learning_rate": 1.8768075279492384e-05,
|
| 59 |
+
"loss": 0.3725,
|
| 60 |
+
"step": 4000
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.9425019425019425,
|
| 64 |
+
"grad_norm": 3.3194727897644043,
|
| 65 |
+
"learning_rate": 1.790477834851297e-05,
|
| 66 |
+
"loss": 0.3419,
|
| 67 |
+
"step": 5000
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 2.0,
|
| 71 |
+
"eval_bleu": 0.18133955287882247,
|
| 72 |
+
"eval_brevity_penalty": 0.19999763034774826,
|
| 73 |
+
"eval_length_ratio": 0.38322255328902133,
|
| 74 |
+
"eval_loss": 0.24950601160526276,
|
| 75 |
+
"eval_precision": [
|
| 76 |
+
0.948350991114149,
|
| 77 |
+
0.9136170304639765,
|
| 78 |
+
0.8917892531319569,
|
| 79 |
+
0.874732452172378
|
| 80 |
+
],
|
| 81 |
+
"eval_reference_length": 122164,
|
| 82 |
+
"eval_rouge1": 0.5697347111060407,
|
| 83 |
+
"eval_rouge2": 0.49477362898706856,
|
| 84 |
+
"eval_rougeL": 0.5685279495432232,
|
| 85 |
+
"eval_rougeLsum": 0.5678546442295118,
|
| 86 |
+
"eval_runtime": 18.6979,
|
| 87 |
+
"eval_samples_per_second": 26.687,
|
| 88 |
+
"eval_steps_per_second": 1.711,
|
| 89 |
+
"eval_translation_length": 46816,
|
| 90 |
+
"step": 5148
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 2.331002331002331,
|
| 94 |
+
"grad_norm": 4.120768070220947,
|
| 95 |
+
"learning_rate": 1.7041481417533562e-05,
|
| 96 |
+
"loss": 0.2611,
|
| 97 |
+
"step": 6000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 2.7195027195027195,
|
| 101 |
+
"grad_norm": 3.0822534561157227,
|
| 102 |
+
"learning_rate": 1.6178184486554153e-05,
|
| 103 |
+
"loss": 0.2509,
|
| 104 |
+
"step": 7000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 3.0,
|
| 108 |
+
"eval_bleu": 0.18212865975864842,
|
| 109 |
+
"eval_brevity_penalty": 0.19960755503568892,
|
| 110 |
+
"eval_length_ratio": 0.38293605317442125,
|
| 111 |
+
"eval_loss": 0.21081040799617767,
|
| 112 |
+
"eval_precision": [
|
| 113 |
+
0.9456189478634488,
|
| 114 |
+
0.9178514325223629,
|
| 115 |
+
0.900399711683376,
|
| 116 |
+
0.8869137002031623
|
| 117 |
+
],
|
| 118 |
+
"eval_reference_length": 122164,
|
| 119 |
+
"eval_rouge1": 0.5793831879084954,
|
| 120 |
+
"eval_rouge2": 0.5102329869808768,
|
| 121 |
+
"eval_rougeL": 0.5768346718094006,
|
| 122 |
+
"eval_rougeLsum": 0.5769243674226936,
|
| 123 |
+
"eval_runtime": 18.8084,
|
| 124 |
+
"eval_samples_per_second": 26.531,
|
| 125 |
+
"eval_steps_per_second": 1.701,
|
| 126 |
+
"eval_translation_length": 46781,
|
| 127 |
+
"step": 7722
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 3.108003108003108,
|
| 131 |
+
"grad_norm": 0.7676980495452881,
|
| 132 |
+
"learning_rate": 1.5314887555574744e-05,
|
| 133 |
+
"loss": 0.225,
|
| 134 |
+
"step": 8000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 3.4965034965034967,
|
| 138 |
+
"grad_norm": 1.1135072708129883,
|
| 139 |
+
"learning_rate": 1.4451590624595331e-05,
|
| 140 |
+
"loss": 0.1849,
|
| 141 |
+
"step": 9000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 3.885003885003885,
|
| 145 |
+
"grad_norm": 1.668112874031067,
|
| 146 |
+
"learning_rate": 1.358829369361592e-05,
|
| 147 |
+
"loss": 0.186,
|
| 148 |
+
"step": 10000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 4.0,
|
| 152 |
+
"eval_bleu": 0.18147618243887598,
|
| 153 |
+
"eval_brevity_penalty": 0.19618265451046102,
|
| 154 |
+
"eval_length_ratio": 0.38041485216594084,
|
| 155 |
+
"eval_loss": 0.19280420243740082,
|
| 156 |
+
"eval_precision": [
|
| 157 |
+
0.9548124717577948,
|
| 158 |
+
0.929503632487928,
|
| 159 |
+
0.91412864211105,
|
| 160 |
+
0.9025257915332622
|
| 161 |
+
],
|
| 162 |
+
"eval_reference_length": 122164,
|
| 163 |
+
"eval_rouge1": 0.5898421422204958,
|
| 164 |
+
"eval_rouge2": 0.5267291316375613,
|
| 165 |
+
"eval_rougeL": 0.5879837779143762,
|
| 166 |
+
"eval_rougeLsum": 0.5875372471265154,
|
| 167 |
+
"eval_runtime": 18.7555,
|
| 168 |
+
"eval_samples_per_second": 26.606,
|
| 169 |
+
"eval_steps_per_second": 1.706,
|
| 170 |
+
"eval_translation_length": 46473,
|
| 171 |
+
"step": 10296
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 4.273504273504273,
|
| 175 |
+
"grad_norm": 0.5733640789985657,
|
| 176 |
+
"learning_rate": 1.272499676263651e-05,
|
| 177 |
+
"loss": 0.1611,
|
| 178 |
+
"step": 11000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 4.662004662004662,
|
| 182 |
+
"grad_norm": 2.5280842781066895,
|
| 183 |
+
"learning_rate": 1.1861699831657098e-05,
|
| 184 |
+
"loss": 0.1524,
|
| 185 |
+
"step": 12000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 5.0,
|
| 189 |
+
"eval_bleu": 0.18384348522783198,
|
| 190 |
+
"eval_brevity_penalty": 0.1991397001931579,
|
| 191 |
+
"eval_length_ratio": 0.3825922530369012,
|
| 192 |
+
"eval_loss": 0.18783439695835114,
|
| 193 |
+
"eval_precision": [
|
| 194 |
+
0.9514538180106549,
|
| 195 |
+
0.9271626297577855,
|
| 196 |
+
0.912769725191841,
|
| 197 |
+
0.9021042394235445
|
| 198 |
+
],
|
| 199 |
+
"eval_reference_length": 122164,
|
| 200 |
+
"eval_rouge1": 0.5917741970317836,
|
| 201 |
+
"eval_rouge2": 0.5290776452842172,
|
| 202 |
+
"eval_rougeL": 0.5892081711516338,
|
| 203 |
+
"eval_rougeLsum": 0.5896535407048884,
|
| 204 |
+
"eval_runtime": 18.7508,
|
| 205 |
+
"eval_samples_per_second": 26.612,
|
| 206 |
+
"eval_steps_per_second": 1.707,
|
| 207 |
+
"eval_translation_length": 46739,
|
| 208 |
+
"step": 12870
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 5.05050505050505,
|
| 212 |
+
"grad_norm": 1.9729936122894287,
|
| 213 |
+
"learning_rate": 1.0998402900677689e-05,
|
| 214 |
+
"loss": 0.1446,
|
| 215 |
+
"step": 13000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 5.439005439005439,
|
| 219 |
+
"grad_norm": 0.6781195402145386,
|
| 220 |
+
"learning_rate": 1.0135105969698278e-05,
|
| 221 |
+
"loss": 0.1251,
|
| 222 |
+
"step": 14000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 5.827505827505828,
|
| 226 |
+
"grad_norm": 1.632392406463623,
|
| 227 |
+
"learning_rate": 9.271809038718868e-06,
|
| 228 |
+
"loss": 0.1275,
|
| 229 |
+
"step": 15000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 6.0,
|
| 233 |
+
"eval_bleu": 0.18328986521134696,
|
| 234 |
+
"eval_brevity_penalty": 0.19751536151098467,
|
| 235 |
+
"eval_length_ratio": 0.38139713827314103,
|
| 236 |
+
"eval_loss": 0.18608906865119934,
|
| 237 |
+
"eval_precision": [
|
| 238 |
+
0.954156203721589,
|
| 239 |
+
0.9320952835510045,
|
| 240 |
+
0.9182366487553459,
|
| 241 |
+
0.9080627993613625
|
| 242 |
+
],
|
| 243 |
+
"eval_reference_length": 122164,
|
| 244 |
+
"eval_rouge1": 0.5954654576993901,
|
| 245 |
+
"eval_rouge2": 0.5350441771248188,
|
| 246 |
+
"eval_rougeL": 0.5942791204166212,
|
| 247 |
+
"eval_rougeLsum": 0.5941621149236106,
|
| 248 |
+
"eval_runtime": 18.7908,
|
| 249 |
+
"eval_samples_per_second": 26.556,
|
| 250 |
+
"eval_steps_per_second": 1.703,
|
| 251 |
+
"eval_translation_length": 46593,
|
| 252 |
+
"step": 15444
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 6.216006216006216,
|
| 256 |
+
"grad_norm": 0.6160931587219238,
|
| 257 |
+
"learning_rate": 8.408512107739457e-06,
|
| 258 |
+
"loss": 0.112,
|
| 259 |
+
"step": 16000
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 6.6045066045066045,
|
| 263 |
+
"grad_norm": 1.4462038278579712,
|
| 264 |
+
"learning_rate": 7.545215176760047e-06,
|
| 265 |
+
"loss": 0.1069,
|
| 266 |
+
"step": 17000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 6.993006993006993,
|
| 270 |
+
"grad_norm": 0.9217690229415894,
|
| 271 |
+
"learning_rate": 6.681918245780636e-06,
|
| 272 |
+
"loss": 0.1059,
|
| 273 |
+
"step": 18000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 7.0,
|
| 277 |
+
"eval_bleu": 0.18324051669709965,
|
| 278 |
+
"eval_brevity_penalty": 0.19728198284540724,
|
| 279 |
+
"eval_length_ratio": 0.381225238204381,
|
| 280 |
+
"eval_loss": 0.1790783554315567,
|
| 281 |
+
"eval_precision": [
|
| 282 |
+
0.9542214205960663,
|
| 283 |
+
0.9325852451544289,
|
| 284 |
+
0.9193180322113486,
|
| 285 |
+
0.9097726012201885
|
| 286 |
+
],
|
| 287 |
+
"eval_reference_length": 122164,
|
| 288 |
+
"eval_rouge1": 0.5936401806586702,
|
| 289 |
+
"eval_rouge2": 0.5338457392086675,
|
| 290 |
+
"eval_rougeL": 0.5924260179841824,
|
| 291 |
+
"eval_rougeLsum": 0.5921117263026487,
|
| 292 |
+
"eval_runtime": 18.7792,
|
| 293 |
+
"eval_samples_per_second": 26.572,
|
| 294 |
+
"eval_steps_per_second": 1.704,
|
| 295 |
+
"eval_translation_length": 46572,
|
| 296 |
+
"step": 18018
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.381507381507381,
|
| 300 |
+
"grad_norm": 0.705454409122467,
|
| 301 |
+
"learning_rate": 5.818621314801226e-06,
|
| 302 |
+
"loss": 0.0966,
|
| 303 |
+
"step": 19000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 7.77000777000777,
|
| 307 |
+
"grad_norm": 1.3365228176116943,
|
| 308 |
+
"learning_rate": 4.955324383821816e-06,
|
| 309 |
+
"loss": 0.0879,
|
| 310 |
+
"step": 20000
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 8.0,
|
| 314 |
+
"eval_bleu": 0.1839110607571646,
|
| 315 |
+
"eval_brevity_penalty": 0.19762651721368382,
|
| 316 |
+
"eval_length_ratio": 0.38147899544874103,
|
| 317 |
+
"eval_loss": 0.17582955956459045,
|
| 318 |
+
"eval_precision": [
|
| 319 |
+
0.9530931485097526,
|
| 320 |
+
0.9341271906992885,
|
| 321 |
+
0.9222672952527136,
|
| 322 |
+
0.9133818117323638
|
| 323 |
+
],
|
| 324 |
+
"eval_reference_length": 122164,
|
| 325 |
+
"eval_rouge1": 0.5973063198680042,
|
| 326 |
+
"eval_rouge2": 0.5411237678060034,
|
| 327 |
+
"eval_rougeL": 0.5952889632065425,
|
| 328 |
+
"eval_rougeLsum": 0.5956406860763515,
|
| 329 |
+
"eval_runtime": 18.7833,
|
| 330 |
+
"eval_samples_per_second": 26.566,
|
| 331 |
+
"eval_steps_per_second": 1.704,
|
| 332 |
+
"eval_translation_length": 46603,
|
| 333 |
+
"step": 20592
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 8.158508158508159,
|
| 337 |
+
"grad_norm": 1.9379663467407227,
|
| 338 |
+
"learning_rate": 4.092027452842406e-06,
|
| 339 |
+
"loss": 0.0878,
|
| 340 |
+
"step": 21000
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 8.547008547008547,
|
| 344 |
+
"grad_norm": 0.6134539842605591,
|
| 345 |
+
"learning_rate": 3.228730521862995e-06,
|
| 346 |
+
"loss": 0.0829,
|
| 347 |
+
"step": 22000
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 8.935508935508935,
|
| 351 |
+
"grad_norm": 0.8216579556465149,
|
| 352 |
+
"learning_rate": 2.3654335908835845e-06,
|
| 353 |
+
"loss": 0.0815,
|
| 354 |
+
"step": 23000
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 9.0,
|
| 358 |
+
"eval_bleu": 0.18513737863673796,
|
| 359 |
+
"eval_brevity_penalty": 0.1993067616515038,
|
| 360 |
+
"eval_length_ratio": 0.3827150388003012,
|
| 361 |
+
"eval_loss": 0.17754927277565002,
|
| 362 |
+
"eval_precision": [
|
| 363 |
+
0.9512768960944518,
|
| 364 |
+
0.9323100205383201,
|
| 365 |
+
0.9204913016872104,
|
| 366 |
+
0.9120136111540756
|
| 367 |
+
],
|
| 368 |
+
"eval_reference_length": 122164,
|
| 369 |
+
"eval_rouge1": 0.5975489794777522,
|
| 370 |
+
"eval_rouge2": 0.5411326394192295,
|
| 371 |
+
"eval_rougeL": 0.5950592400363829,
|
| 372 |
+
"eval_rougeLsum": 0.5959819950305132,
|
| 373 |
+
"eval_runtime": 18.7891,
|
| 374 |
+
"eval_samples_per_second": 26.558,
|
| 375 |
+
"eval_steps_per_second": 1.703,
|
| 376 |
+
"eval_translation_length": 46754,
|
| 377 |
+
"step": 23166
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 9.324009324009324,
|
| 381 |
+
"grad_norm": 2.143796443939209,
|
| 382 |
+
"learning_rate": 1.5021366599041743e-06,
|
| 383 |
+
"loss": 0.0807,
|
| 384 |
+
"step": 24000
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 9.712509712509712,
|
| 388 |
+
"grad_norm": 1.8318359851837158,
|
| 389 |
+
"learning_rate": 6.388397289247637e-07,
|
| 390 |
+
"loss": 0.0789,
|
| 391 |
+
"step": 25000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 10.0,
|
| 395 |
+
"eval_bleu": 0.1853679051126517,
|
| 396 |
+
"eval_brevity_penalty": 0.19985272444672936,
|
| 397 |
+
"eval_length_ratio": 0.3831161389607413,
|
| 398 |
+
"eval_loss": 0.1786118447780609,
|
| 399 |
+
"eval_precision": [
|
| 400 |
+
0.9501527679849582,
|
| 401 |
+
0.9309346924671734,
|
| 402 |
+
0.919069970527235,
|
| 403 |
+
0.910409217322209
|
| 404 |
+
],
|
| 405 |
+
"eval_reference_length": 122164,
|
| 406 |
+
"eval_rouge1": 0.5953647536190028,
|
| 407 |
+
"eval_rouge2": 0.5386002688888253,
|
| 408 |
+
"eval_rougeL": 0.5936608393631321,
|
| 409 |
+
"eval_rougeLsum": 0.5933789948328584,
|
| 410 |
+
"eval_runtime": 18.7839,
|
| 411 |
+
"eval_samples_per_second": 26.565,
|
| 412 |
+
"eval_steps_per_second": 1.704,
|
| 413 |
+
"eval_translation_length": 46803,
|
| 414 |
+
"step": 25740
|
| 415 |
+
}
|
| 416 |
+
],
|
| 417 |
+
"logging_steps": 1000,
|
| 418 |
+
"max_steps": 25740,
|
| 419 |
+
"num_input_tokens_seen": 0,
|
| 420 |
+
"num_train_epochs": 10,
|
| 421 |
+
"save_steps": 500,
|
| 422 |
+
"stateful_callbacks": {
|
| 423 |
+
"TrainerControl": {
|
| 424 |
+
"args": {
|
| 425 |
+
"should_epoch_stop": false,
|
| 426 |
+
"should_evaluate": false,
|
| 427 |
+
"should_log": false,
|
| 428 |
+
"should_save": true,
|
| 429 |
+
"should_training_stop": true
|
| 430 |
+
},
|
| 431 |
+
"attributes": {}
|
| 432 |
+
}
|
| 433 |
+
},
|
| 434 |
+
"total_flos": 3.122366168448e+16,
|
| 435 |
+
"train_batch_size": 4,
|
| 436 |
+
"trial_name": null,
|
| 437 |
+
"trial_params": null
|
| 438 |
+
}
|
checkpoint-25740/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2226bbd434dc747522e0462a07f6683302d39a2395301b5fa3230230b5d764b9
|
| 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": 0,
|
| 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.1751331117.gra5
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94385bc3fd94903b3c26df276e3bae907e2496a1905d22bbe99c2f9c2ac5acc1
|
| 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:d3204e3f87f0aa024b6c789452cc8948835758c7e1220ddba095a680a7121e2f
|
| 3 |
+
size 891558696
|
results.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"rouge1": 0.7358212004336553,
|
| 3 |
+
"rouge2": 0.6797578699194634,
|
| 4 |
+
"rougeL": 0.7329282653086759,
|
| 5 |
+
"rougeLsum": 0.7328429342715225,
|
| 6 |
+
"bleu": 0.6547934677850505,
|
| 7 |
+
"precision": [
|
| 8 |
+
0.6776294623616751,
|
| 9 |
+
0.657776167658108,
|
| 10 |
+
0.6464586934084122,
|
| 11 |
+
0.6379786693130931
|
| 12 |
+
],
|
| 13 |
+
"brevity_penalty": 1.0,
|
| 14 |
+
"length_ratio": 1.447014646786271,
|
| 15 |
+
"translation_length": 329478,
|
| 16 |
+
"reference_length": 227695
|
| 17 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2226bbd434dc747522e0462a07f6683302d39a2395301b5fa3230230b5d764b9
|
| 3 |
+
size 5560
|