Merge pull request #6 from facebookresearch/HF_paths_update

Hf paths update

README.md

@@ -63,8 +63,8 @@ As part of the seamless communication project, we've extended the stopes library
 ## SeamlessM4T models
 | Model Name         | #params | checkpoint                                                                              | metrics                                                                              |
 | ------------------ | ------- | --------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------ |
-| SeamlessM4T-Large  | 2.3B    | [model](https://dl.fbaipublicfiles.com/seamlessM4T/models/large/seamlessM4T_large.pt)   | [metrics](https://dl.fbaipublicfiles.com/seamlessM4T/metrics/seamlessM4T_large.zip)  |
-| SeamlessM4T-Medium | 1.2B    | [model](https://dl.fbaipublicfiles.com/seamlessM4T/models/medium/seamlessM4T_medium.pt) | [metrics](https://dl.fbaipublicfiles.com/seamlessM4T/metrics/seamlessM4T_medium.zip) |
+| SeamlessM4T-Large  | 2.3B    | [model](https://huggingface.co/facebook/seamless-m4t-large/resolve/main/multitask_unity_large.pt)   | [metrics](https://dl.fbaipublicfiles.com/seamlessM4T/metrics/seamlessM4T_large.zip)  |
+| SeamlessM4T-Medium | 1.2B    | [model](https://huggingface.co/facebook/seamless-m4t-medium/resolve/main/multitask_unity_medium.pt) | [metrics](https://dl.fbaipublicfiles.com/seamlessM4T/metrics/seamlessM4T_medium.zip) |
 
 We provide the extensive evaluation results of seamlessM4T-Large and SeamlessM4T-Medium reported in the paper (as averages) in the `metrics` files above.
 

docs/m4t/on_device_README.md

@@ -6,8 +6,8 @@ This folder contains an example to run an exported small model covering most tas
 ## Overview
 | Model   | Num Params | Disk Size | Supported Tasks         | Supported Languages|
 |---------|----------------------|-------------|------------|-------------------------|
-| [UnitY-Small](https://dl.fbaipublicfiles.com/seamless_aug/models/small_unity/unity_on_device.ptl) | 281M | 862MB        | S2ST, S2TT, ASR |eng, fra, hin, por, spa|
-| [UnitY-Small-S2T](https://dl.fbaipublicfiles.com/seamless_aug/models/small_unity/unity_on_device_s2t.ptl) | 235M | 637MB        | S2TT, ASR    |eng, fra,hin,  por, spa|
+| [UnitY-Small](https://huggingface.co/facebook/seamless-m4t-unity-small/resolve/main/unity_on_device.ptl) | 281M | 862MB        | S2ST, S2TT, ASR |eng, fra, hin, por, spa|
+| [UnitY-Small-S2T](https://huggingface.co/facebook/seamless-m4t-unity-small-s2t/resolve/main/unity_on_device_s2t.ptl) | 235M | 637MB        | S2TT, ASR    |eng, fra,hin,  por, spa|
 
 UnitY-Small-S2T is a pruned version of UnitY-Small without 2nd pass unit decoding.
 

src/seamless_communication/assets/cards/seamlessM4T_large.yaml

@@ -7,7 +7,7 @@
 name: seamlessM4T_large
 base: unity_nllb-100
 model_arch: base
-checkpoint: "https://dl.fbaipublicfiles.com/seamless_aug/models/large_unity/multitask_unity_large.pt"
+checkpoint: "https://huggingface.co/facebook/seamless-m4t-large/resolve/main/multitask_unity_large.pt"
 num_units: 10000
 unit_langs:
   - arb

src/seamless_communication/assets/cards/seamlessM4T_medium.yaml

@@ -7,7 +7,7 @@
 name: seamlessM4T_medium
 base: unity_nllb-200
 model_arch: medium
-checkpoint: "https://dl.fbaipublicfiles.com/seamless_aug/models/medium_unity/multitask_unity_medium.pt"
+checkpoint: "https://huggingface.co/facebook/seamless-m4t-medium/resolve/main/multitask_unity_medium.pt"
 num_units: 10000
 unit_langs:
   - arb

src/seamless_communication/assets/cards/unity_nllb-100.yaml

@@ -5,7 +5,7 @@
 # LICENSE file in the root directory of this source tree.
 
 model_type: unity
-tokenizer: "https://dl.fbaipublicfiles.com/seamless_aug/models/large_unity/spm_256k_nllb100.model"
+tokenizer: "https://dl.fbaipublicfiles.com/seamlessM4T/models/large/spm_256k_nllb100.model"
 default_lang: eng
 langs:
   - afr

src/seamless_communication/assets/cards/vocoder_36langs.yaml

@@ -7,7 +7,7 @@
 name: vocoder_36langs
 model_type: vocoder_code_hifigan
 model_arch: base
-checkpoint: "https://dl.fbaipublicfiles.com/seamless_aug/models/vocoder/vocoder_36langs.pt"
+checkpoint: "https://huggingface.co/facebook/seamless-m4t-vocoder/resolve/main/vocoder_36langs.pt"
 model_config: {
     "lang_spkr_idx_map": {
         "multilingual": {