ibm-granite/granite-speech-4.1-2b-GGUF

Same weights, converted (fewer layers) vs ibm-granite/granite-speech-4.1-2b. ibm-granite/granite-speech-4.1-2b-GGUF matches part of ibm-granite/granite-speech-4.1-2b (363 of its 954 tensors; the rest of ibm-granite/granite-speech-4.1-2b has no counterpart). On the shared part: Same weights stored in another dtype (BF16/I64 → BF16/F32).

Tensors changed: Embeddings 0/1, Norms 0/81, MLP 0/120, Attention 0/160, lm_head 0/1

On the Hub · Downloads on Model Pulse