Hugging Face's logo

ysdede
/
Phi-4-mm-inst-asr-turkish-unf

Text Generation
Transformers
TensorBoard
Safetensors
Turkish
phi4mm
Generated from Trainer
custom_code
Model card Files Metrics Community
ysdede commited on
Commit
00ab220
·
verified ·
1 Parent(s): 49975e8

Training in progress, epoch 1

Browse files
Files changed (10) hide show
  1. config.json +221 -0
  2. model-00001-of-00003.safetensors +3 -0
  3. model-00002-of-00003.safetensors +3 -0
  4. model-00003-of-00003.safetensors +3 -0
  5. model.safetensors.index.json +0 -0
  6. runs/Mar06_22-12-11_d8fde9f7ee2f/events.out.tfevents.1741299132.d8fde9f7ee2f.8580.0 +3 -0
  7. runs/Mar06_22-22-23_d8fde9f7ee2f/events.out.tfevents.1741299743.d8fde9f7ee2f.8580.1 +3 -0
  8. runs/Mar06_22-26-10_d8fde9f7ee2f/events.out.tfevents.1741299970.d8fde9f7ee2f.8580.2 +3 -0
  9. runs/Mar07_11-03-04_361a5899d46c/events.out.tfevents.1741345385.361a5899d46c.287.0 +3 -0
  10. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,221 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/Phi-4-multimodal-instruct",
3
+ "architectures": [
4
+ "Phi4MMForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "audio_processor": {
9
+ "config": {
10
+ "activation": "swish",
11
+ "activation_checkpointing": {
12
+ "interval": 1,
13
+ "module": "transformer",
14
+ "offload": false
15
+ },
16
+ "attention_dim": 1024,
17
+ "attention_heads": 16,
18
+ "batch_norm": false,
19
+ "bias_in_glu": true,
20
+ "causal": true,
21
+ "chunk_size": -1,
22
+ "cnn_layer_norm": true,
23
+ "conv_activation": "swish",
24
+ "conv_glu_type": "swish",
25
+ "depthwise_multiplier": 1,
26
+ "depthwise_seperable_out_channel": 1024,
27
+ "dropout_rate": 0.0,
28
+ "encoder_embedding_config": {
29
+ "input_size": 80
30
+ },
31
+ "ext_pw_kernel_size": 1,
32
+ "ext_pw_out_channel": 1024,
33
+ "input_layer": "nemo_conv",
34
+ "input_size": 80,
35
+ "kernel_size": 3,
36
+ "left_chunk": 18,
37
+ "linear_units": 1536,
38
+ "nemo_conv_settings": {
39
+ "conv_channels": 1024
40
+ },
41
+ "num_blocks": 24,
42
+ "relative_attention_bias_args": {
43
+ "t5_bias_max_distance": 500,
44
+ "type": "t5"
45
+ },
46
+ "time_reduction": 8
47
+ },
48
+ "name": "cascades"
49
+ },
50
+ "auto_map": {
51
+ "AutoConfig": "configuration_phi4mm.Phi4MMConfig",
52
+ "AutoModelForCausalLM": "modeling_phi4mm.Phi4MMForCausalLM",
53
+ "AutoTokenizer": "microsoft/Phi-4-multimodal-instruct--Xenova/gpt-4o"
54
+ },
55
+ "bos_token_id": 199999,
56
+ "embd_layer": {
57
+ "audio_embd_layer": {
58
+ "compression_rate": 8,
59
+ "downsample_rate": 1,
60
+ "embedding_cls": "audio",
61
+ "enable_gradient_checkpointing": true,
62
+ "projection_cls": "mlp",
63
+ "use_conv_downsample": false,
64
+ "use_qformer": false
65
+ },
66
+ "embedding_cls": "image_audio",
67
+ "image_embd_layer": {
68
+ "crop_size": 448,
69
+ "embedding_cls": "tune_image",
70
+ "enable_gradient_checkpointing": true,
71
+ "hd_transform_order": "sub_glb",
72
+ "image_token_compression_cls": "avg_pool_2d",
73
+ "projection_cls": "mlp",
74
+ "use_hd_transform": true,
75
+ "with_learnable_separator": true
76
+ }
77
+ },
78
+ "embd_pdrop": 0.0,
79
+ "eos_token_id": 199999,
80
+ "full_attn_mod": 1,
81
+ "hidden_act": "silu",
82
+ "hidden_size": 3072,
83
+ "img_processor": null,
84
+ "initializer_range": 0.02,
85
+ "intermediate_size": 8192,
86
+ "interpolate_factor": 1,
87
+ "lm_head_bias": false,
88
+ "max_position_embeddings": 131072,
89
+ "mlp_bias": false,
90
+ "model_type": "phi4mm",
91
+ "num_attention_heads": 24,
92
+ "num_hidden_layers": 32,
93
+ "num_key_value_heads": 8,
94
+ "original_max_position_embeddings": 4096,
95
+ "pad_token_id": 199999,
96
+ "partial_rotary_factor": 0.75,
97
+ "resid_pdrop": 0.0,
98
+ "rms_norm_eps": 1e-05,
99
+ "rope_scaling": {
100
+ "long_factor": [
101
+ 1,
102
+ 1.118320672,
103
+ 1.250641126,
104
+ 1.398617824,
105
+ 1.564103225,
106
+ 1.74916897,
107
+ 1.956131817,
108
+ 2.187582649,
109
+ 2.446418898,
110
+ 2.735880826,
111
+ 3.059592084,
112
+ 3.421605075,
113
+ 3.826451687,
114
+ 4.279200023,
115
+ 4.785517845,
116
+ 5.351743533,
117
+ 5.984965424,
118
+ 6.693110555,
119
+ 7.485043894,
120
+ 8.370679318,
121
+ 9.36110372,
122
+ 10.4687158,
123
+ 11.70738129,
124
+ 13.09260651,
125
+ 14.64173252,
126
+ 16.37415215,
127
+ 18.31155283,
128
+ 20.47818807,
129
+ 22.90118105,
130
+ 25.61086418,
131
+ 28.64115884,
132
+ 32.03,
133
+ 32.1,
134
+ 32.13,
135
+ 32.23,
136
+ 32.6,
137
+ 32.61,
138
+ 32.64,
139
+ 32.66,
140
+ 32.7,
141
+ 32.71,
142
+ 32.93,
143
+ 32.97,
144
+ 33.28,
145
+ 33.49,
146
+ 33.5,
147
+ 44.16,
148
+ 47.77
149
+ ],
150
+ "short_factor": [
151
+ 1.0,
152
+ 1.0,
153
+ 1.0,
154
+ 1.0,
155
+ 1.0,
156
+ 1.0,
157
+ 1.0,
158
+ 1.0,
159
+ 1.0,
160
+ 1.0,
161
+ 1.0,
162
+ 1.0,
163
+ 1.0,
164
+ 1.0,
165
+ 1.0,
166
+ 1.0,
167
+ 1.0,
168
+ 1.0,
169
+ 1.0,
170
+ 1.0,
171
+ 1.0,
172
+ 1.0,
173
+ 1.0,
174
+ 1.0,
175
+ 1.0,
176
+ 1.0,
177
+ 1.0,
178
+ 1.0,
179
+ 1.0,
180
+ 1.0,
181
+ 1.0,
182
+ 1.0,
183
+ 1.0,
184
+ 1.0,
185
+ 1.0,
186
+ 1.0,
187
+ 1.0,
188
+ 1.0,
189
+ 1.0,
190
+ 1.0,
191
+ 1.0,
192
+ 1.0,
193
+ 1.0,
194
+ 1.0,
195
+ 1.0,
196
+ 1.0,
197
+ 1.0,
198
+ 1.0
199
+ ],
200
+ "type": "longrope"
201
+ },
202
+ "rope_theta": 10000.0,
203
+ "sliding_window": 262144,
204
+ "speech_lora": {
205
+ "dp": 0.01,
206
+ "layer": "((layers.*self_attn\\.(qkv|o)_proj)|(layers.*mlp\\.(gate_up|down)_proj))",
207
+ "lora_alpha": 640,
208
+ "r": 320
209
+ },
210
+ "tie_word_embeddings": true,
211
+ "torch_dtype": "bfloat16",
212
+ "transformers_version": "4.48.3",
213
+ "use_cache": true,
214
+ "vision_lora": {
215
+ "dp": 0.0,
216
+ "layer": "layers.*((self_attn\\.(qkv_proj|o_proj))|(mlp\\.(gate_up|down)_proj))",
217
+ "lora_alpha": 512,
218
+ "r": 256
219
+ },
220
+ "vocab_size": 200064
221
+ }
model-00001-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:74498dc1c50f440d682b1c648b37e683e82c26fd13e09484d6c744ee37c339a1
3
+ size 4997504848
model-00002-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d457a44e5136a1934ca0076244fd09abc41cd92d1066a5aa1ec08c398b9e15e5
3
+ size 4952333128
model-00003-of-00003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:49275ac99190ff6d42c6ccc7158ed7ab62cbad4859b9e8bfc27da717aab179a7
3
+ size 1199389232
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
runs/Mar06_22-12-11_d8fde9f7ee2f/events.out.tfevents.1741299132.d8fde9f7ee2f.8580.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63684208c34cda6131362713b6dd3c7bd9439e942389b07fb8d7c03554529186
3
+ size 9272
runs/Mar06_22-22-23_d8fde9f7ee2f/events.out.tfevents.1741299743.d8fde9f7ee2f.8580.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d33a26c4f7c492e406230df033477a2f405b4000564eb3e97fbad5a6970bba4
3
+ size 9272
runs/Mar06_22-26-10_d8fde9f7ee2f/events.out.tfevents.1741299970.d8fde9f7ee2f.8580.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e906a91d4dc131d6810f1ac32cbbe615991982bbe8a44d2cb21dc6ed8fc4f89
3
+ size 8280
runs/Mar07_11-03-04_361a5899d46c/events.out.tfevents.1741345385.361a5899d46c.287.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7eb0126ae48cf9aa4d775f08d80c08b8c8386d920db19eac51cd21cc5826977
3
+ size 29069
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:90c5437c9e86dead385c5440c56814fe0c08eb6a362aea68b256737b936eed30
3
+ size 5432