feat: update weights and init xclbin

.gitattributes

@@ -33,3 +33,9 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+decoder_mv.xclbin filter=lfs diff=lfs merge=lfs -text
+dequant.xclbin filter=lfs diff=lfs merge=lfs -text
+encoder_attn.xclbin filter=lfs diff=lfs merge=lfs -text
+encoder_mm.xclbin filter=lfs diff=lfs merge=lfs -text
+lm_head.xclbin filter=lfs diff=lfs merge=lfs -text
+model.q4nx filter=lfs diff=lfs merge=lfs -text

config.json

@@ -0,0 +1,49 @@
+{
+  "_name_or_path": "/raid/yoach/tmp_whisper_turbo",
+  "activation_dropout": 0.0,
+  "activation_function": "gelu",
+  "apply_spec_augment": false,
+  "architectures": [
+    "WhisperForConditionalGeneration"
+  ],
+  "attention_dropout": 0.0,
+  "begin_suppress_tokens": [
+    220,
+    50256
+  ],
+  "bos_token_id": 50257,
+  "classifier_proj_size": 256,
+  "d_model": 1280,
+  "decoder_attention_heads": 20,
+  "decoder_ffn_dim": 5120,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 4,
+  "decoder_start_token_id": 50258,
+  "dropout": 0.0,
+  "encoder_attention_heads": 20,
+  "encoder_ffn_dim": 5120,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 32,
+  "eos_token_id": 50257,
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
+  "max_source_positions": 1500,
+  "max_target_positions": 448,
+  "median_filter_width": 7,
+  "model_type": "whisper",
+  "num_hidden_layers": 32,
+  "num_mel_bins": 128,
+  "pad_token_id": 50257,
+  "scale_embedding": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.46.0.dev0",
+  "use_cache": true,
+  "use_weighted_layer_sum": false,
+  "vocab_size": 51866
+}

decoder_mv.xclbin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f711363051bcd6a688031ce917a53b8f4e088b1f4eda53bf352e9581443d749
+size 137051

dequant.xclbin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9bfd2ee346dbdcd074aaf595e86b2f8e7505bb383276506ed6a1e0280f2cb6a
+size 114667

encoder_attn.xclbin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adc23d80f7ecd77e7567bf2b14ed143630f8286bb3e7b6a2a85223ac9e6f4ae2
+size 383083

encoder_mm.xclbin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b40727f1fe20be8981863491b992247d1f64cfbef40949569be2f5d557d4e639
+size 163211

lm_head.xclbin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2053faba39db087c3045e1dda8c27702001a7af58229d4016d9e8c2c3e45027a
+size 153355

model.q4nx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fb97604bf5762ee26efa696cfc9eb70724110358c7b4ca628a7973bf8a16291
+size 650175128