Add the en-fr config.

README.md

@@ -3,8 +3,8 @@ Delayed Streams Modeling (DSM) is a flexible formulation for streaming, multimod
 
 ## Speech To Text
 
-### Leaderboard model
+### English only model
-The leaderboard model handles english only, it has ~2.6B parameters.
+The main model handles english only, it has ~2.6B parameters.
 
 #### PyTorch implementation
 [[Hugging Face]](https://huggingface.co/kyutai/stt-2.6B-en)
@@ -55,6 +55,19 @@ The script simulates some real-time processing of the audio. Faster processing
 can be triggered by setting the real-time factor, e.g. `--rtf 500` will process
 the data as fast as possible.
 
+### English + French model
+This model has ~1B parameters and supports both English and French.
+
+#### Rust implementation
+[[Hugging Face]](https://huggingface.co/kyutai/stt-1B-en_fr-candle)
+
+The only difference with the en only model is the config file used when
+launching the server.
+```bash
+moshi-server worker --config configs/config-stt-enfr-hf.toml
+```
+
+
 ## Text To Speech
 
 We're in the process of open-sourcing our TTS models. Check back for updates!

configs/config-stt-enfr-hf.toml

@@ -0,0 +1,46 @@
+static_dir = "./static/"
+log_dir = "$HOME/tmp/tts-logs"
+instance_name = "tts"
+authorized_ids = ["open_token"]
+
+[modules.asr]
+path = "/api/asr-streaming"
+type = "BatchedAsr"
+lm_model_file = "hf://kyutai/stt-1B-en_fr-candle/model.safetensors"
+text_tokenizer_file = "hf://kyutai/stt-1B-en_fr-candle/tokenizer_en_fr_audio_8000.model"
+audio_tokenizer_file = "hf://kyutai/stt-1B-en_fr-candle/mimi-pytorch-e351c8d8@125.safetensors"
+asr_delay_in_tokens = 6
+batch_size = 64
+conditioning_learnt_padding = true
+temperature = 0.0
+
+[modules.asr.model]
+audio_vocab_size = 2049
+text_in_vocab_size = 8001
+text_out_vocab_size = 8000
+audio_codebooks = 32
+
+[modules.asr.model.transformer]
+d_model = 2048
+num_heads = 16
+num_layers = 16
+dim_feedforward = 8192
+causal = true
+norm_first = true
+bias_ff = false
+bias_attn = false
+context = 750
+max_period = 100000
+use_conv_block = false
+use_conv_bias = true
+gating = "silu"
+norm = "RmsNorm"
+positional_embedding = "Rope"
+conv_layout = false
+conv_kernel_size = 3
+kv_repeat = 1
+max_seq_len = 40960
+
+[modules.asr.model.extra_heads]
+num_heads = 4
+dim = 6