[GH-ISSUE #12281] Model request: ERNIE-4.5-21B-A3B-Thinking - enable thinking, tool use, and full quantizations #54676

Open
opened 2026-04-29 06:52:25 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @zytoh0 on GitHub (Sep 14, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12281

Hi Ollama team

Requesting adding baidu/ERNIE-4.5-21B-A3B-Thinking to Ollama.

Requests

  • Thinking

  • Tools

  • Publish standard set: Q2_K, Q3_K_S/M/L, Q4_0/1, Q4_K_S/M, Q5_0/1, Q5_K_S/M, Q6_K, Q8_0 (+ BF16/FP16 if feasible).

  • Defaults: num_ctx=131072; align chat template with HF format.

Thank you :)

Originally created by @zytoh0 on GitHub (Sep 14, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12281 Hi Ollama team Requesting adding [baidu/ERNIE-4.5-21B-A3B-Thinking](https://huggingface.co/baidu/ERNIE-4.5-21B-A3B-Thinking) to Ollama. - Model: [https://huggingface.co/baidu/ERNIE-4.5-21B-A3B-Thinking](https://huggingface.co/baidu/ERNIE-4.5-21B-A3B-Thinking) - License: Apache-2.0 - Highlights: 21B MoE (~3B active), 128K context, reasoning “thinking” variant, function/tool calling, HF Transformers weights (BF16/F32). **Requests** - Thinking - Tools - Publish standard set: Q2_K, Q3_K_S/M/L, Q4_0/1, Q4_K_S/M, Q5_0/1, Q5_K_S/M, Q6_K, Q8_0 (+ BF16/FP16 if feasible). - Defaults: num_ctx=131072; align chat template with HF format. Thank you :)
GiteaMirror added the model label 2026-04-29 06:52:25 -05:00
Author
Owner

@itzpingcat commented on GitHub (Dec 7, 2025):

upvote

<!-- gh-comment-id:3622811050 --> @itzpingcat commented on GitHub (Dec 7, 2025): upvote
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#54676