hudsongouge/minicpm5-1B-GLM-5.2-Agentic-v7

Fix BOS token: set add_bos_token=True, remove bos_token from template - fixes degenerate output in LM Studio

Fix BOS token: set add_bos_token=True, remove bos_token from template - fixes degenerate output in LM Studio

Fix BOS token: set add_bos_token=True, remove bos_token from template - fixes degenerate output in LM Studio

Fix BOS token: set add_bos_token=True, remove bos_token from template - fixes degenerate output in LM Studio

Fix BOS token: set add_bos_token=True, remove bos_token from template - fixes degenerate output in LM Studio

Fix BOS token: set add_bos_token=True, remove bos_token from template - fixes degenerate output in LM Studio

Use exact official OpenBMB chat template (enable_thinking is defined) for LM Studio compatibility

Use exact official OpenBMB chat template (enable_thinking is defined) for LM Studio compatibility

Use exact official OpenBMB chat template (enable_thinking is defined) for LM Studio compatibility

Use exact official OpenBMB chat template (enable_thinking is defined) for LM Studio compatibility

Use exact official OpenBMB chat template (enable_thinking is defined) for LM Studio compatibility

Use exact official OpenBMB chat template (enable_thinking is defined) for LM Studio compatibility

Fix pre-tokenizer: use llama-bpe instead of minicpm5 for LM Studio compatibility

Fix pre-tokenizer: use llama-bpe instead of minicpm5 for LM Studio compatibility

Fix pre-tokenizer: use llama-bpe instead of minicpm5 for LM Studio compatibility

Fix pre-tokenizer: use llama-bpe instead of minicpm5 for LM Studio compatibility

Fix pre-tokenizer: use llama-bpe instead of minicpm5 for LM Studio compatibility

Fix pre-tokenizer: use llama-bpe instead of minicpm5 for LM Studio compatibility

Re-convert GGUF from safetensors with fixed chat template (proper alignment)

Re-convert GGUF from safetensors with fixed chat template (proper alignment)

Re-convert GGUF from safetensors with fixed chat template (proper alignment)

Re-convert GGUF from safetensors with fixed chat template (proper alignment)

Re-convert GGUF from safetensors with fixed chat template (proper alignment)

Re-convert GGUF from safetensors with fixed chat template (proper alignment)

Fix chat template: default enable_thinking=true (if false -> skip, else -> thinking)

Add quantization recommendations and chat template notes

Fix chat template: default enable_thinking=true so model generates reasoning + response

Fix chat template: default enable_thinking=true so model generates reasoning + response

Fix chat template: default enable_thinking=true so model generates reasoning + response

Fix chat template: default enable_thinking=true so model generates reasoning + response

Fix chat template: default enable_thinking=true so model generates reasoning + response

Fix chat template: default enable_thinking=true so model generates reasoning + response

Fix EOS token id (1 -> 130073)

Fix EOS token id (1 -> 130073)

Fix EOS token id (1 -> 130073)

Fix EOS token id (1 -> 130073)

Fix EOS token id (1 -> 130073)

Fix EOS token id (1 -> 130073)

Update model card with eval data

Add q2_k GGUF

Add q3_k_m GGUF

Add q4_k_m GGUF

Add q5_k_m GGUF

Add q8_0 GGUF

Add f16 GGUF

Upload v7

initial commit