mazesmazes commited on
Commit
6e07bb2
·
verified ·
1 Parent(s): ab38b9e

Training in progress, step 1000

Browse files
adapter_config.json CHANGED
@@ -3,7 +3,7 @@
3
  "alpha_pattern": {},
4
  "arrow_config": null,
5
  "auto_mapping": null,
6
- "base_model_name_or_path": "checkpoint-18000",
7
  "bias": "none",
8
  "corda_config": null,
9
  "ensure_weight_tying": false,
@@ -29,8 +29,8 @@
29
  "rank_pattern": {},
30
  "revision": null,
31
  "target_modules": [
32
- "v_proj",
33
- "q_proj"
34
  ],
35
  "target_parameters": null,
36
  "task_type": "CAUSAL_LM",
 
3
  "alpha_pattern": {},
4
  "arrow_config": null,
5
  "auto_mapping": null,
6
+ "base_model_name_or_path": "mazesmazes/tiny-audio",
7
  "bias": "none",
8
  "corda_config": null,
9
  "ensure_weight_tying": false,
 
29
  "rank_pattern": {},
30
  "revision": null,
31
  "target_modules": [
32
+ "q_proj",
33
+ "v_proj"
34
  ],
35
  "target_parameters": null,
36
  "task_type": "CAUSAL_LM",
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8041780442726f6dfe4977f7481dc095fc6641f80f8efd6b511ce6698c79c44c
3
  size 36715216
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3d9525bb79a88033e2abf0745be479d053bc7e472082abe550fc21970f0171f
3
  size 36715216
config.json CHANGED
@@ -277,6 +277,7 @@
277
  ],
278
  "encoder_dim": 1280,
279
  "freeze_projector": false,
 
280
  "inference_warmup_tokens": 10,
281
  "label_smoothing": 0.0,
282
  "length_penalty": 1.0,
@@ -288,12 +289,6 @@
288
  "q_proj",
289
  "v_proj"
290
  ],
291
- "mask_feature_length": 10,
292
- "mask_feature_min_masks": 2,
293
- "mask_feature_prob": 0.05,
294
- "mask_time_length": 15,
295
- "mask_time_min_masks": 2,
296
- "mask_time_prob": 0.1,
297
  "max_new_tokens": 128,
298
  "min_new_tokens": 0,
299
  "model_dtype": "bfloat16",
@@ -302,6 +297,8 @@
302
  "num_beams": 1,
303
  "num_experts": 4,
304
  "num_experts_per_tok": 2,
 
 
305
  "pipeline_tag": "automatic-speech-recognition",
306
  "pretrained_model_path": "mazesmazes/tiny-audio",
307
  "projector_dropout": 0.0,
@@ -381,6 +378,7 @@
381
  "vocab_size": 151670
382
  },
383
  "text_model_id": "Qwen/Qwen3-0.6B",
 
384
  "transformers_version": "5.0.0.dev0",
385
  "use_cache": false,
386
  "use_lora": true,
 
277
  ],
278
  "encoder_dim": 1280,
279
  "freeze_projector": false,
280
+ "freq_mask_length": 10,
281
  "inference_warmup_tokens": 10,
282
  "label_smoothing": 0.0,
283
  "length_penalty": 1.0,
 
289
  "q_proj",
290
  "v_proj"
291
  ],
 
 
 
 
 
 
292
  "max_new_tokens": 128,
293
  "min_new_tokens": 0,
294
  "model_dtype": "bfloat16",
 
297
  "num_beams": 1,
298
  "num_experts": 4,
299
  "num_experts_per_tok": 2,
300
+ "num_freq_masks": 2,
301
+ "num_time_masks": 2,
302
  "pipeline_tag": "automatic-speech-recognition",
303
  "pretrained_model_path": "mazesmazes/tiny-audio",
304
  "projector_dropout": 0.0,
 
378
  "vocab_size": 151670
379
  },
380
  "text_model_id": "Qwen/Qwen3-0.6B",
381
+ "time_mask_length": 15,
382
  "transformers_version": "5.0.0.dev0",
383
  "use_cache": false,
384
  "use_lora": true,
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1d884484522ff44b7cc38a1a0fd035e7fb7cb3d7954709726db5761e42ec88a3
3
  size 25172384
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:82b3a651cbcecc1ac7e3a970b54119ae5a9c12b53c5e6f3fee991ab69340d5f3
3
  size 25172384
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6cb17eae71944b178ffca38182449645e5a12a962efd2d3eb01b9a84d725d5ed
3
  size 5201
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:043f6ed7f53a2db67ac5f2249ac181fc40cb4bef3a20d41c2b1f48eae3176980
3
  size 5201