Upload 9 files
Browse files- config.json +82 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- preprocessor_config.json +10 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +507 -0
- training_args.bin +3 -0
- vocab.json +1 -0
config.json
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/w2v-bert-2.0",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_act": "relu",
|
| 5 |
+
"adapter_kernel_size": 3,
|
| 6 |
+
"adapter_stride": 2,
|
| 7 |
+
"add_adapter": true,
|
| 8 |
+
"apply_spec_augment": false,
|
| 9 |
+
"architectures": [
|
| 10 |
+
"Wav2Vec2BertForCTC"
|
| 11 |
+
],
|
| 12 |
+
"attention_dropout": 0.0,
|
| 13 |
+
"bos_token_id": 1,
|
| 14 |
+
"classifier_proj_size": 768,
|
| 15 |
+
"codevector_dim": 768,
|
| 16 |
+
"conformer_conv_dropout": 0.1,
|
| 17 |
+
"contrastive_logits_temperature": 0.1,
|
| 18 |
+
"conv_depthwise_kernel_size": 31,
|
| 19 |
+
"ctc_loss_reduction": "mean",
|
| 20 |
+
"ctc_zero_infinity": true,
|
| 21 |
+
"diversity_loss_weight": 0.1,
|
| 22 |
+
"eos_token_id": 2,
|
| 23 |
+
"feat_proj_dropout": 0.0,
|
| 24 |
+
"feat_quantizer_dropout": 0.0,
|
| 25 |
+
"feature_projection_input_dim": 160,
|
| 26 |
+
"final_dropout": 0.1,
|
| 27 |
+
"hidden_act": "swish",
|
| 28 |
+
"hidden_dropout": 0.0,
|
| 29 |
+
"hidden_size": 1024,
|
| 30 |
+
"initializer_range": 0.02,
|
| 31 |
+
"intermediate_size": 4096,
|
| 32 |
+
"layer_norm_eps": 1e-05,
|
| 33 |
+
"layerdrop": 0.0,
|
| 34 |
+
"left_max_position_embeddings": 64,
|
| 35 |
+
"mask_feature_length": 10,
|
| 36 |
+
"mask_feature_min_masks": 0,
|
| 37 |
+
"mask_feature_prob": 0.0,
|
| 38 |
+
"mask_time_length": 10,
|
| 39 |
+
"mask_time_min_masks": 2,
|
| 40 |
+
"mask_time_prob": 0.0,
|
| 41 |
+
"max_source_positions": 5000,
|
| 42 |
+
"model_type": "wav2vec2-bert",
|
| 43 |
+
"num_adapter_layers": 1,
|
| 44 |
+
"num_attention_heads": 16,
|
| 45 |
+
"num_codevector_groups": 2,
|
| 46 |
+
"num_codevectors_per_group": 320,
|
| 47 |
+
"num_hidden_layers": 24,
|
| 48 |
+
"num_negatives": 100,
|
| 49 |
+
"output_hidden_size": 1024,
|
| 50 |
+
"pad_token_id": 64,
|
| 51 |
+
"position_embeddings_type": "relative_key",
|
| 52 |
+
"proj_codevector_dim": 768,
|
| 53 |
+
"right_max_position_embeddings": 8,
|
| 54 |
+
"rotary_embedding_base": 10000,
|
| 55 |
+
"tdnn_dilation": [
|
| 56 |
+
1,
|
| 57 |
+
2,
|
| 58 |
+
3,
|
| 59 |
+
1,
|
| 60 |
+
1
|
| 61 |
+
],
|
| 62 |
+
"tdnn_dim": [
|
| 63 |
+
512,
|
| 64 |
+
512,
|
| 65 |
+
512,
|
| 66 |
+
512,
|
| 67 |
+
1500
|
| 68 |
+
],
|
| 69 |
+
"tdnn_kernel": [
|
| 70 |
+
5,
|
| 71 |
+
3,
|
| 72 |
+
3,
|
| 73 |
+
1,
|
| 74 |
+
1
|
| 75 |
+
],
|
| 76 |
+
"torch_dtype": "float32",
|
| 77 |
+
"transformers_version": "4.42.4",
|
| 78 |
+
"use_intermediate_ffn_before_adapter": false,
|
| 79 |
+
"use_weighted_layer_sum": false,
|
| 80 |
+
"vocab_size": 67,
|
| 81 |
+
"xvector_output_dim": 512
|
| 82 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a837d41c8036372cc0627a0caac22e196da7378c1b2a88714a19ecf4fba2b85b
|
| 3 |
+
size 2423089260
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8ea1a799f5b20d1e126ff8e2a37ebb2c3bb39d7c793d8a8e6559229234fcb50
|
| 3 |
+
size 4846639530
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"feature_extractor_type": "SeamlessM4TFeatureExtractor",
|
| 3 |
+
"feature_size": 80,
|
| 4 |
+
"num_mel_bins": 80,
|
| 5 |
+
"padding_side": "right",
|
| 6 |
+
"padding_value": 0.0,
|
| 7 |
+
"return_attention_mask": true,
|
| 8 |
+
"sampling_rate": 16000,
|
| 9 |
+
"stride": 2
|
| 10 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ffddd8c01ad7e3933f495e5a745ea09d4a45713f5f13035d7b1547b81929f153
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:713d9a5941ba462e0a0a288525ba91540a6b9134cb37cb8af5e06f9237c31c57
|
| 3 |
+
size 1256
|
trainer_state.json
ADDED
|
@@ -0,0 +1,507 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.23266415805525312,
|
| 3 |
+
"best_model_checkpoint": "results/punjabi_model/20240813_012221/checkpoint-5000",
|
| 4 |
+
"epoch": 1.9260709757154553,
|
| 5 |
+
"eval_steps": 5000,
|
| 6 |
+
"global_step": 30000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03210118292859092,
|
| 13 |
+
"grad_norm": Infinity,
|
| 14 |
+
"learning_rate": 1.24e-05,
|
| 15 |
+
"loss": 2.1834,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06420236585718184,
|
| 20 |
+
"grad_norm": 9.310239791870117,
|
| 21 |
+
"learning_rate": 2.4900000000000002e-05,
|
| 22 |
+
"loss": 0.676,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09630354878577276,
|
| 27 |
+
"grad_norm": 8.342071533203125,
|
| 28 |
+
"learning_rate": 3.74e-05,
|
| 29 |
+
"loss": 0.577,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12840473171436367,
|
| 34 |
+
"grad_norm": 8.718728065490723,
|
| 35 |
+
"learning_rate": 4.99e-05,
|
| 36 |
+
"loss": 0.5662,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1605059146429546,
|
| 41 |
+
"grad_norm": 8.425736427307129,
|
| 42 |
+
"learning_rate": 4.98390243902439e-05,
|
| 43 |
+
"loss": 0.5139,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1926070975715455,
|
| 48 |
+
"grad_norm": 7.612177848815918,
|
| 49 |
+
"learning_rate": 4.967642276422764e-05,
|
| 50 |
+
"loss": 0.4737,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.22470828050013644,
|
| 55 |
+
"grad_norm": 6.828492164611816,
|
| 56 |
+
"learning_rate": 4.951382113821139e-05,
|
| 57 |
+
"loss": 0.4696,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.25680946342872735,
|
| 62 |
+
"grad_norm": 25.135059356689453,
|
| 63 |
+
"learning_rate": 4.9351219512195126e-05,
|
| 64 |
+
"loss": 0.4487,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2889106463573183,
|
| 69 |
+
"grad_norm": 6.97437047958374,
|
| 70 |
+
"learning_rate": 4.9188617886178865e-05,
|
| 71 |
+
"loss": 0.4259,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3210118292859092,
|
| 76 |
+
"grad_norm": 4.6233744621276855,
|
| 77 |
+
"learning_rate": 4.9026016260162604e-05,
|
| 78 |
+
"loss": 0.3788,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3210118292859092,
|
| 83 |
+
"eval_loss": 0.4430150091648102,
|
| 84 |
+
"eval_runtime": 597.4312,
|
| 85 |
+
"eval_samples_per_second": 11.601,
|
| 86 |
+
"eval_steps_per_second": 11.601,
|
| 87 |
+
"eval_wer": 0.23266415805525312,
|
| 88 |
+
"step": 5000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.3531130122145001,
|
| 92 |
+
"grad_norm": 4.593179225921631,
|
| 93 |
+
"learning_rate": 4.886341463414634e-05,
|
| 94 |
+
"loss": 0.4035,
|
| 95 |
+
"step": 5500
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.385214195143091,
|
| 99 |
+
"grad_norm": 6.695476055145264,
|
| 100 |
+
"learning_rate": 4.870081300813009e-05,
|
| 101 |
+
"loss": 0.3882,
|
| 102 |
+
"step": 6000
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.41731537807168195,
|
| 106 |
+
"grad_norm": 14.509389877319336,
|
| 107 |
+
"learning_rate": 4.853853658536586e-05,
|
| 108 |
+
"loss": 0.3574,
|
| 109 |
+
"step": 6500
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.4494165610002729,
|
| 113 |
+
"grad_norm": 38.900367736816406,
|
| 114 |
+
"learning_rate": 4.8375934959349596e-05,
|
| 115 |
+
"loss": 0.3723,
|
| 116 |
+
"step": 7000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.48151774392886376,
|
| 120 |
+
"grad_norm": 3.945736885070801,
|
| 121 |
+
"learning_rate": 4.8213333333333335e-05,
|
| 122 |
+
"loss": 0.3586,
|
| 123 |
+
"step": 7500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.5136189268574547,
|
| 127 |
+
"grad_norm": 77.0046615600586,
|
| 128 |
+
"learning_rate": 4.8050731707317074e-05,
|
| 129 |
+
"loss": 0.3679,
|
| 130 |
+
"step": 8000
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.5457201097860456,
|
| 134 |
+
"grad_norm": 16.720748901367188,
|
| 135 |
+
"learning_rate": 4.788813008130081e-05,
|
| 136 |
+
"loss": 0.3352,
|
| 137 |
+
"step": 8500
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.5778212927146366,
|
| 141 |
+
"grad_norm": 17.356231689453125,
|
| 142 |
+
"learning_rate": 4.772552845528456e-05,
|
| 143 |
+
"loss": 0.341,
|
| 144 |
+
"step": 9000
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.6099224756432274,
|
| 148 |
+
"grad_norm": 9.849752426147461,
|
| 149 |
+
"learning_rate": 4.75629268292683e-05,
|
| 150 |
+
"loss": 0.3354,
|
| 151 |
+
"step": 9500
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.6420236585718184,
|
| 155 |
+
"grad_norm": 9.619921684265137,
|
| 156 |
+
"learning_rate": 4.7400650406504066e-05,
|
| 157 |
+
"loss": 0.3288,
|
| 158 |
+
"step": 10000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.6420236585718184,
|
| 162 |
+
"eval_loss": 0.35230785608291626,
|
| 163 |
+
"eval_runtime": 596.403,
|
| 164 |
+
"eval_samples_per_second": 11.621,
|
| 165 |
+
"eval_steps_per_second": 11.621,
|
| 166 |
+
"eval_wer": 0.18700615692192107,
|
| 167 |
+
"step": 10000
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.6741248415004093,
|
| 171 |
+
"grad_norm": 7.539575099945068,
|
| 172 |
+
"learning_rate": 4.723804878048781e-05,
|
| 173 |
+
"loss": 0.3314,
|
| 174 |
+
"step": 10500
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.7062260244290002,
|
| 178 |
+
"grad_norm": 13.222583770751953,
|
| 179 |
+
"learning_rate": 4.7075447154471544e-05,
|
| 180 |
+
"loss": 0.3383,
|
| 181 |
+
"step": 11000
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.7383272073575912,
|
| 185 |
+
"grad_norm": 2.821357011795044,
|
| 186 |
+
"learning_rate": 4.691284552845528e-05,
|
| 187 |
+
"loss": 0.3319,
|
| 188 |
+
"step": 11500
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.770428390286182,
|
| 192 |
+
"grad_norm": 7.006532669067383,
|
| 193 |
+
"learning_rate": 4.675024390243903e-05,
|
| 194 |
+
"loss": 0.3218,
|
| 195 |
+
"step": 12000
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.8025295732147729,
|
| 199 |
+
"grad_norm": 6.498229503631592,
|
| 200 |
+
"learning_rate": 4.658764227642277e-05,
|
| 201 |
+
"loss": 0.3175,
|
| 202 |
+
"step": 12500
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.8346307561433639,
|
| 206 |
+
"grad_norm": 8.40401840209961,
|
| 207 |
+
"learning_rate": 4.6425365853658536e-05,
|
| 208 |
+
"loss": 0.3083,
|
| 209 |
+
"step": 13000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.8667319390719548,
|
| 213 |
+
"grad_norm": 1.3094803094863892,
|
| 214 |
+
"learning_rate": 4.626276422764228e-05,
|
| 215 |
+
"loss": 0.3109,
|
| 216 |
+
"step": 13500
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.8988331220005458,
|
| 220 |
+
"grad_norm": 3.827925443649292,
|
| 221 |
+
"learning_rate": 4.610016260162602e-05,
|
| 222 |
+
"loss": 0.3203,
|
| 223 |
+
"step": 14000
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.9309343049291366,
|
| 227 |
+
"grad_norm": 2.329725503921509,
|
| 228 |
+
"learning_rate": 4.593756097560975e-05,
|
| 229 |
+
"loss": 0.2979,
|
| 230 |
+
"step": 14500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.9630354878577275,
|
| 234 |
+
"grad_norm": 3.5614166259765625,
|
| 235 |
+
"learning_rate": 4.57749593495935e-05,
|
| 236 |
+
"loss": 0.3024,
|
| 237 |
+
"step": 15000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.9630354878577275,
|
| 241 |
+
"eval_loss": 0.30602961778640747,
|
| 242 |
+
"eval_runtime": 592.7853,
|
| 243 |
+
"eval_samples_per_second": 11.692,
|
| 244 |
+
"eval_steps_per_second": 11.692,
|
| 245 |
+
"eval_wer": 0.16886647692905213,
|
| 246 |
+
"step": 15000
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.9951366707863185,
|
| 250 |
+
"grad_norm": 87.19066619873047,
|
| 251 |
+
"learning_rate": 4.561235772357724e-05,
|
| 252 |
+
"loss": 0.3088,
|
| 253 |
+
"step": 15500
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.0272378537149094,
|
| 257 |
+
"grad_norm": 2.4488048553466797,
|
| 258 |
+
"learning_rate": 4.544975609756098e-05,
|
| 259 |
+
"loss": 0.2761,
|
| 260 |
+
"step": 16000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.0593390366435003,
|
| 264 |
+
"grad_norm": 2.8186426162719727,
|
| 265 |
+
"learning_rate": 4.5287154471544715e-05,
|
| 266 |
+
"loss": 0.2708,
|
| 267 |
+
"step": 16500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.0914402195720911,
|
| 271 |
+
"grad_norm": 1.2367055416107178,
|
| 272 |
+
"learning_rate": 4.5124552845528454e-05,
|
| 273 |
+
"loss": 0.2814,
|
| 274 |
+
"step": 17000
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 1.1235414025006822,
|
| 278 |
+
"grad_norm": 1.350964903831482,
|
| 279 |
+
"learning_rate": 4.496227642276423e-05,
|
| 280 |
+
"loss": 0.2943,
|
| 281 |
+
"step": 17500
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.1556425854292731,
|
| 285 |
+
"grad_norm": 3.4485456943511963,
|
| 286 |
+
"learning_rate": 4.479967479674797e-05,
|
| 287 |
+
"loss": 0.2814,
|
| 288 |
+
"step": 18000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 1.187743768357864,
|
| 292 |
+
"grad_norm": 2.5614452362060547,
|
| 293 |
+
"learning_rate": 4.463707317073171e-05,
|
| 294 |
+
"loss": 0.2624,
|
| 295 |
+
"step": 18500
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 1.2198449512864549,
|
| 299 |
+
"grad_norm": 5.0949625968933105,
|
| 300 |
+
"learning_rate": 4.447447154471545e-05,
|
| 301 |
+
"loss": 0.2808,
|
| 302 |
+
"step": 19000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.2519461342150457,
|
| 306 |
+
"grad_norm": 1.0048319101333618,
|
| 307 |
+
"learning_rate": 4.431186991869919e-05,
|
| 308 |
+
"loss": 0.2534,
|
| 309 |
+
"step": 19500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.2840473171436368,
|
| 313 |
+
"grad_norm": 1.9653562307357788,
|
| 314 |
+
"learning_rate": 4.4149268292682924e-05,
|
| 315 |
+
"loss": 0.2465,
|
| 316 |
+
"step": 20000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.2840473171436368,
|
| 320 |
+
"eval_loss": 0.28473955392837524,
|
| 321 |
+
"eval_runtime": 595.99,
|
| 322 |
+
"eval_samples_per_second": 11.629,
|
| 323 |
+
"eval_steps_per_second": 11.629,
|
| 324 |
+
"eval_wer": 0.14723766227978022,
|
| 325 |
+
"step": 20000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.3161485000722277,
|
| 329 |
+
"grad_norm": 0.9238675832748413,
|
| 330 |
+
"learning_rate": 4.3987317073170736e-05,
|
| 331 |
+
"loss": 0.2812,
|
| 332 |
+
"step": 20500
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.3482496830008186,
|
| 336 |
+
"grad_norm": 1.7164336442947388,
|
| 337 |
+
"learning_rate": 4.3824715447154475e-05,
|
| 338 |
+
"loss": 0.2716,
|
| 339 |
+
"step": 21000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.3803508659294095,
|
| 343 |
+
"grad_norm": 1.9793908596038818,
|
| 344 |
+
"learning_rate": 4.3662113821138214e-05,
|
| 345 |
+
"loss": 0.2779,
|
| 346 |
+
"step": 21500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.4124520488580004,
|
| 350 |
+
"grad_norm": 2.866766929626465,
|
| 351 |
+
"learning_rate": 4.349951219512195e-05,
|
| 352 |
+
"loss": 0.2467,
|
| 353 |
+
"step": 22000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.4445532317865912,
|
| 357 |
+
"grad_norm": 1.6922576427459717,
|
| 358 |
+
"learning_rate": 4.333691056910569e-05,
|
| 359 |
+
"loss": 0.2616,
|
| 360 |
+
"step": 22500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.4766544147151823,
|
| 364 |
+
"grad_norm": 2.3735485076904297,
|
| 365 |
+
"learning_rate": 4.317430894308943e-05,
|
| 366 |
+
"loss": 0.273,
|
| 367 |
+
"step": 23000
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.5087555976437732,
|
| 371 |
+
"grad_norm": 1.882097601890564,
|
| 372 |
+
"learning_rate": 4.3012032520325206e-05,
|
| 373 |
+
"loss": 0.2541,
|
| 374 |
+
"step": 23500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.540856780572364,
|
| 378 |
+
"grad_norm": 1.2024834156036377,
|
| 379 |
+
"learning_rate": 4.2849430894308945e-05,
|
| 380 |
+
"loss": 0.2679,
|
| 381 |
+
"step": 24000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.572957963500955,
|
| 385 |
+
"grad_norm": 1.467214584350586,
|
| 386 |
+
"learning_rate": 4.2686829268292684e-05,
|
| 387 |
+
"loss": 0.2788,
|
| 388 |
+
"step": 24500
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.605059146429546,
|
| 392 |
+
"grad_norm": 0.9247603416442871,
|
| 393 |
+
"learning_rate": 4.252422764227643e-05,
|
| 394 |
+
"loss": 0.2575,
|
| 395 |
+
"step": 25000
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.605059146429546,
|
| 399 |
+
"eval_loss": 0.26871952414512634,
|
| 400 |
+
"eval_runtime": 596.0677,
|
| 401 |
+
"eval_samples_per_second": 11.628,
|
| 402 |
+
"eval_steps_per_second": 11.628,
|
| 403 |
+
"eval_wer": 0.13321745331372287,
|
| 404 |
+
"step": 25000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.637160329358137,
|
| 408 |
+
"grad_norm": 1.6064761877059937,
|
| 409 |
+
"learning_rate": 4.236162601626016e-05,
|
| 410 |
+
"loss": 0.2325,
|
| 411 |
+
"step": 25500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.6692615122867278,
|
| 415 |
+
"grad_norm": 1.4238929748535156,
|
| 416 |
+
"learning_rate": 4.21990243902439e-05,
|
| 417 |
+
"loss": 0.2463,
|
| 418 |
+
"step": 26000
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 1.7013626952153187,
|
| 422 |
+
"grad_norm": 0.7233666181564331,
|
| 423 |
+
"learning_rate": 4.2036422764227646e-05,
|
| 424 |
+
"loss": 0.2537,
|
| 425 |
+
"step": 26500
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.7334638781439096,
|
| 429 |
+
"grad_norm": 0.9025558233261108,
|
| 430 |
+
"learning_rate": 4.1873821138211385e-05,
|
| 431 |
+
"loss": 0.2524,
|
| 432 |
+
"step": 27000
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 1.7655650610725004,
|
| 436 |
+
"grad_norm": 2.8646883964538574,
|
| 437 |
+
"learning_rate": 4.1711219512195124e-05,
|
| 438 |
+
"loss": 0.2536,
|
| 439 |
+
"step": 27500
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.7976662440010913,
|
| 443 |
+
"grad_norm": 1.5429887771606445,
|
| 444 |
+
"learning_rate": 4.154861788617886e-05,
|
| 445 |
+
"loss": 0.2664,
|
| 446 |
+
"step": 28000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.8297674269296822,
|
| 450 |
+
"grad_norm": 2.8908677101135254,
|
| 451 |
+
"learning_rate": 4.138634146341464e-05,
|
| 452 |
+
"loss": 0.2636,
|
| 453 |
+
"step": 28500
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.8618686098582733,
|
| 457 |
+
"grad_norm": 2.929805278778076,
|
| 458 |
+
"learning_rate": 4.122373983739838e-05,
|
| 459 |
+
"loss": 0.257,
|
| 460 |
+
"step": 29000
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.8939697927868642,
|
| 464 |
+
"grad_norm": 1.2041730880737305,
|
| 465 |
+
"learning_rate": 4.1061138211382116e-05,
|
| 466 |
+
"loss": 0.2459,
|
| 467 |
+
"step": 29500
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.9260709757154553,
|
| 471 |
+
"grad_norm": 1.7089420557022095,
|
| 472 |
+
"learning_rate": 4.0898536585365855e-05,
|
| 473 |
+
"loss": 0.2608,
|
| 474 |
+
"step": 30000
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.9260709757154553,
|
| 478 |
+
"eval_loss": 0.2512572705745697,
|
| 479 |
+
"eval_runtime": 594.6622,
|
| 480 |
+
"eval_samples_per_second": 11.655,
|
| 481 |
+
"eval_steps_per_second": 11.655,
|
| 482 |
+
"eval_wer": 0.1310080925002706,
|
| 483 |
+
"step": 30000
|
| 484 |
+
}
|
| 485 |
+
],
|
| 486 |
+
"logging_steps": 500,
|
| 487 |
+
"max_steps": 155750,
|
| 488 |
+
"num_input_tokens_seen": 0,
|
| 489 |
+
"num_train_epochs": 10,
|
| 490 |
+
"save_steps": 5000,
|
| 491 |
+
"stateful_callbacks": {
|
| 492 |
+
"TrainerControl": {
|
| 493 |
+
"args": {
|
| 494 |
+
"should_epoch_stop": false,
|
| 495 |
+
"should_evaluate": false,
|
| 496 |
+
"should_log": false,
|
| 497 |
+
"should_save": true,
|
| 498 |
+
"should_training_stop": false
|
| 499 |
+
},
|
| 500 |
+
"attributes": {}
|
| 501 |
+
}
|
| 502 |
+
},
|
| 503 |
+
"total_flos": 2.520783816961326e+19,
|
| 504 |
+
"train_batch_size": 1,
|
| 505 |
+
"trial_name": null,
|
| 506 |
+
"trial_params": null
|
| 507 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17e701a2666b059164627cbdcf780a7bc4685100c06c549201703aadc5788ce9
|
| 3 |
+
size 5368
|
vocab.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"ਂ": 1, "ਅ": 2, "ਆ": 3, "ਇ": 4, "ਈ": 5, "ਉ": 6, "ਊ": 7, "ਏ": 8, "ਐ": 9, "ਓ": 10, "ਔ": 11, "ਕ": 12, "ਖ": 13, "ਗ": 14, "ਘ": 15, "ਚ": 16, "ਛ": 17, "ਜ": 18, "ਝ": 19, "ਟ": 20, "ਠ": 21, "ਡ": 22, "ਢ": 23, "ਣ": 24, "ਤ": 25, "ਥ": 26, "ਦ": 27, "ਧ": 28, "ਨ": 29, "ਪ": 30, "ਫ": 31, "ਬ": 32, "ਭ": 33, "ਮ": 34, "ਯ": 35, "ਰ": 36, "ਲ": 37, "ਲ਼": 38, "ਵ": 39, "ਸ਼": 40, "ਸ": 41, "ਹ": 42, "਼": 43, "ਾ": 44, "ਿ": 45, "ੀ": 46, "ੁ": 47, "ੂ": 48, "ੇ": 49, "ੈ": 50, "ੋ": 51, "ੌ": 52, "੍": 53, "ਖ਼": 54, "ਗ਼": 55, "ਜ਼": 56, "ੜ": 57, "ਫ਼": 58, "ੰ": 59, "ੱ": 60, "ੲ": 61, "ੳ": 62, "|": 0, "[UNK]": 63, "[PAD]": 64}
|