tensorblock/granite-34b-code-instruct-8k-GGUF

Quantization vs ibm-granite/granite-34b-code-instruct-8k. tensorblock/granite-34b-code-instruct-8k-GGUF is a GGUF quantization (Q3_K_M) of the same weights: 707 tensors match exactly and 353 within quantization error.

Tensors changed: Attention 0/352, Norms 0/354, MLP 0/352, Embeddings 0/2

On the Hub · Downloads on Model Pulse