Audio-to-Audio
PEFT
Safetensors
Moshi
Turkish
Hindi
speech-to-speech-translation
simultaneous-translation
mimi
lora
tpu
turkish
hindi
Eval Results (legacy)
Instructions to use tiny-aya-translate/tr-hi-s2st-v0.3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use tiny-aya-translate/tr-hi-s2st-v0.3 with PEFT:
Task type is invalid.
- Moshi
How to use tiny-aya-translate/tr-hi-s2st-v0.3 with Moshi:
# pip install moshi # Run the interactive web server python -m moshi.server --hf-repo "tiny-aya-translate/tr-hi-s2st-v0.3" # Then open https://localhost:8998 in your browser
# pip install moshi import torch from moshi.models import loaders # Load checkpoint info from HuggingFace checkpoint = loaders.CheckpointInfo.from_hf_repo("tiny-aya-translate/tr-hi-s2st-v0.3") # Load the Mimi audio codec mimi = checkpoint.get_mimi(device="cuda") mimi.set_num_codebooks(8) # Encode audio (24kHz, mono) wav = torch.randn(1, 1, 24000 * 10) # [batch, channels, samples] with torch.no_grad(): codes = mimi.encode(wav.cuda()) decoded = mimi.decode(codes) - Notebooks
- Google Colab
- Kaggle
checkpoints/best (main-tree mirror)
Browse files- checkpoints/best/audio_heads.pt +1 -1
- checkpoints/best/depth_decoder.pt +1 -1
- checkpoints/best/metadata.json +4 -4
- checkpoints/best/model_audio_embed.pt +1 -1
- checkpoints/best/peft_adapter/adapter_config.json +6 -6
- checkpoints/best/peft_adapter/adapter_model.safetensors +1 -1
- checkpoints/best/projection.pt +1 -1
- checkpoints/best/text_embed.pt +1 -1
checkpoints/best/audio_heads.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8390277
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ceed2330a825689258e5cae77d0e47033a8940811a036d4b5a3f009a68f1db09
|
| 3 |
size 8390277
|
checkpoints/best/depth_decoder.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1428745845
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:21b18de805cd5c4302624db8317e0f89a0841ec16e92c6ef0e4eaed113db117b
|
| 3 |
size 1428745845
|
checkpoints/best/metadata.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"step":
|
| 3 |
-
"best_val_loss": 2.
|
| 4 |
"config": {
|
| 5 |
"backend": "tpu",
|
| 6 |
"data": {
|
|
@@ -17,7 +17,7 @@
|
|
| 17 |
"num_codebooks": 8,
|
| 18 |
"batch_size": 8,
|
| 19 |
"grad_accum": 1,
|
| 20 |
-
"max_steps":
|
| 21 |
"warmup_steps": 1100,
|
| 22 |
"min_lr_ratio": 0.0,
|
| 23 |
"depth_chunk_size": 100,
|
|
@@ -38,7 +38,7 @@
|
|
| 38 |
"xla_grad_checkpoint": true,
|
| 39 |
"compile_warmup_steps": 1,
|
| 40 |
"debug_input_sharding": false,
|
| 41 |
-
"early_stop_patience":
|
| 42 |
"enable_clip_grad_norm": true,
|
| 43 |
"clip_grad_norm": 1.0,
|
| 44 |
"log_grad_norm": true,
|
|
|
|
| 1 |
{
|
| 2 |
+
"step": 76000,
|
| 3 |
+
"best_val_loss": 2.8199082136154177,
|
| 4 |
"config": {
|
| 5 |
"backend": "tpu",
|
| 6 |
"data": {
|
|
|
|
| 17 |
"num_codebooks": 8,
|
| 18 |
"batch_size": 8,
|
| 19 |
"grad_accum": 1,
|
| 20 |
+
"max_steps": 76250,
|
| 21 |
"warmup_steps": 1100,
|
| 22 |
"min_lr_ratio": 0.0,
|
| 23 |
"depth_chunk_size": 100,
|
|
|
|
| 38 |
"xla_grad_checkpoint": true,
|
| 39 |
"compile_warmup_steps": 1,
|
| 40 |
"debug_input_sharding": false,
|
| 41 |
+
"early_stop_patience": 0,
|
| 42 |
"enable_clip_grad_norm": true,
|
| 43 |
"clip_grad_norm": 1.0,
|
| 44 |
"log_grad_norm": true,
|
checkpoints/best/model_audio_embed.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8394415
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0061a95eef2425d244ecd91c4fd46a8cceb9247548fd4f097eb2bfe95905c6df
|
| 3 |
size 8394415
|
checkpoints/best/peft_adapter/adapter_config.json
CHANGED
|
@@ -30,14 +30,14 @@
|
|
| 30 |
"rank_pattern": {},
|
| 31 |
"revision": null,
|
| 32 |
"target_modules": [
|
| 33 |
-
"up_proj",
|
| 34 |
-
"o_proj",
|
| 35 |
-
"k_proj",
|
| 36 |
"gate_proj",
|
| 37 |
-
"v_proj",
|
| 38 |
-
"embed_tokens",
|
| 39 |
"q_proj",
|
| 40 |
-
"down_proj"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
],
|
| 42 |
"target_parameters": null,
|
| 43 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 30 |
"rank_pattern": {},
|
| 31 |
"revision": null,
|
| 32 |
"target_modules": [
|
|
|
|
|
|
|
|
|
|
| 33 |
"gate_proj",
|
|
|
|
|
|
|
| 34 |
"q_proj",
|
| 35 |
+
"down_proj",
|
| 36 |
+
"o_proj",
|
| 37 |
+
"up_proj",
|
| 38 |
+
"embed_tokens",
|
| 39 |
+
"v_proj",
|
| 40 |
+
"k_proj"
|
| 41 |
],
|
| 42 |
"target_parameters": null,
|
| 43 |
"task_type": "CAUSAL_LM",
|
checkpoints/best/peft_adapter/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1220179520
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a06e3099aee7db5eb80f9d939c47af7cacbbee216a7c4186f5524af35f3a0ff8
|
| 3 |
size 1220179520
|
checkpoints/best/projection.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16778814
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92f655113c8a55c2a0d8e3d275c4fa4fbdd6d2b72ec090225626de92402202f2
|
| 3 |
size 16778814
|
checkpoints/best/text_embed.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1090668924
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c4747e72020bae1a7ca8a75e6704ec3ef8c5ef781bdf06e5a289c0e16e1d300
|
| 3 |
size 1090668924
|