Training in progress, epoch 1

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "vasista22/whisper-hindi-medium",
+  "_name_or_path": "vasista22/whisper-hindi-small",
   "activation_dropout": 0.0,
   "activation_function": "gelu",
   "apply_spec_augment": false,
@@ -13,17 +13,17 @@
   ],
   "bos_token_id": 50257,
   "classifier_proj_size": 256,
-  "d_model": 1024,
-  "decoder_attention_heads": 16,
-  "decoder_ffn_dim": 4096,
+  "d_model": 768,
+  "decoder_attention_heads": 12,
+  "decoder_ffn_dim": 3072,
   "decoder_layerdrop": 0.0,
-  "decoder_layers": 24,
+  "decoder_layers": 12,
   "decoder_start_token_id": 50258,
   "dropout": 0.0,
-  "encoder_attention_heads": 16,
-  "encoder_ffn_dim": 4096,
+  "encoder_attention_heads": 12,
+  "encoder_ffn_dim": 3072,
   "encoder_layerdrop": 0.0,
-  "encoder_layers": 24,
+  "encoder_layers": 12,
   "eos_token_id": 50257,
   "forced_decoder_ids": null,
   "init_std": 0.02,
@@ -39,7 +39,7 @@
   "max_target_positions": 448,
   "median_filter_width": 7,
   "model_type": "whisper",
-  "num_hidden_layers": 24,
+  "num_hidden_layers": 12,
   "num_mel_bins": 80,
   "pad_token_id": 50257,
   "scale_embedding": false,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:317ec66e3ddfcc7e5c61b2f1f8d4e6338b146c62910c14a54267b64049c794ea
-size 3055544304
+oid sha256:fc6c32e299cc86befc94c7d3cd76db62c99553e4317bead1af3d19a0d9908180
+size 966995080

runs/May29_10-55-23_ai-matrix-train/events.out.tfevents.1716980125.ai-matrix-train.118255.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a044ec5794e7d0ca05220756c24a186ce0740d7c0c0f6f6f760abcc0703cdad5
+size 5538

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3e34c368ad52b758cda1131058d403407f69e9e96a8d3366c93386edec2f6ba3
+oid sha256:b408d60db1a12bae19ffb95419de451c9f8ed8a00f3d8b21530290759f4b37b8
 size 5240