Issues

• Training forward pass doesn't match inference architecture — BridgingModule replaces decoder

  davidbrowne17/csm-streaming#50 · Filter by author NeuralNoble opened on Apr 4, 2026

• Loading a finetuned model

  davidbrowne17/csm-streaming#47 · Filter by author kenenbek opened on Sep 14, 2025

• Issue with generating audio

  davidbrowne17/csm-streaming#45 · Filter by author Lor3nzoMartinez opened on Aug 20, 2025
  2

• Getting this error when finetuning using LoRA

  davidbrowne17/csm-streaming#43 · Filter by author arpitjjw opened on Aug 10, 2025
  1

• Full Retraining CSM for different language

  davidbrowne17/csm-streaming#42 · Filter by author Yaman-dot opened on Aug 1, 2025

• Streaming but it's not actually clearing kvbuffer or using sliding windows?

  davidbrowne17/csm-streaming#41 · Filter by author tommedema opened on Jul 5, 2025

• Streaming input

  davidbrowne17/csm-streaming#40 · Filter by author tommedema opened on Jul 5, 2025

• Fine tuning doesn't use context?

  davidbrowne17/csm-streaming#39 · Filter by author tommedema opened on Jul 5, 2025

• vLLM can help to improve the inference speed?

  davidbrowne17/csm-streaming#37 · Filter by author codeview3d opened on Jun 25, 2025

• "python ./lora.py" crashed on "RuntimeError: view size is not compatible with input tensor's size and stride"

  davidbrowne17/csm-streaming#35 · Filter by author jeffDebug opened on Jun 17, 2025
  1

• A silly question about the finetuned model loading

  davidbrowne17/csm-streaming#34 · Filter by author jeffDebug opened on Jun 13, 2025
  1

• Is the decoder parallel on this implementation?

  davidbrowne17/csm-streaming#33 · Filter by author samuel-vitorino opened on May 27, 2025

• "python ./run_csm.py" error: torch._dynamo.exc.RecompileLimitExceeded: cache_size_limit reached

  davidbrowne17/csm-streaming#32 · Filter by author jeffDebug opened on May 21, 2025
  3

• Could not run '_C::ggml_dequantize' with arguments from the 'CUDA' backend

  davidbrowne17/csm-streaming#31 · Filter by author Bewinxed opened on May 20, 2025

• About Training with LoRA on Multiple GPUs

  davidbrowne17/csm-streaming#30 · Filter by author devfed1 opened on May 16, 2025
  2

• Whisper error

  davidbrowne17/csm-streaming#25 · Filter by author fablevi opened on May 8, 2025
  1

• Flat training loss, but model improved

  davidbrowne17/csm-streaming#22 · Filter by author derickdecesare opened on Apr 21, 2025
  4

• Local generation fails to generate persistent voice

  davidbrowne17/csm-streaming#21 · Filter by author orejwan opened on Apr 21, 2025
  5

• Notebook LM Voices

  davidbrowne17/csm-streaming#20 · Filter by author uetuluk opened on Apr 21, 2025
  1

• finetune guide

  davidbrowne17/csm-streaming#19 · Filter by author devops724 opened on Apr 20, 2025
  1

• Failed to initialize codebook_embedding

  davidbrowne17/csm-streaming#15 · Filter by author anpct opened on Apr 18, 2025
  1

• Patch and results on Intel Arc A770 (XPU)

  davidbrowne17/csm-streaming#14 · Filter by author DatCaptainHorse opened on Apr 18, 2025
  1

• Training advice and outcomes

  davidbrowne17/csm-streaming#9 · Filter by author GeEom opened on Apr 11, 2025
  1