Monibee-Fudgekins commited on
Commit
9a1ed42
·
verified ·
1 Parent(s): f4d596f

Fix model card metadata: base_model self-loop -> google/gemma-4-26B-A4B-it, license apache-2.0 -> gemma

Browse files
Files changed (1) hide show
  1. README.md +15 -5
README.md CHANGED
@@ -1,21 +1,31 @@
1
  ---
2
- base_model: Monibee-Fudgekins/gemma-coder
3
  tags:
4
  - text-generation-inference
5
  - transformers
6
  - unsloth
7
  - gemma4
8
- license: apache-2.0
 
 
9
  language:
10
  - en
11
  ---
12
 
13
- # Uploaded finetuned model
 
 
 
 
14
 
15
  - **Developed by:** Monibee-Fudgekins
16
- - **License:** apache-2.0
17
- - **Finetuned from model :** Monibee-Fudgekins/gemma-coder
18
 
19
  This gemma4 model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
20
 
 
 
 
 
21
  [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
 
1
  ---
2
+ base_model: google/gemma-4-26B-A4B-it
3
  tags:
4
  - text-generation-inference
5
  - transformers
6
  - unsloth
7
  - gemma4
8
+ - code
9
+ - coding-assistant
10
+ license: gemma
11
  language:
12
  - en
13
  ---
14
 
15
+ # gemma-coder
16
+
17
+ Coding-focused fine-tune of [`google/gemma-4-26B-A4B-it`](https://huggingface.co/google/gemma-4-26B-A4B-it)
18
+ (**Gemma 4 26B A4B**, an MoE with ~4B active params), produced automatically by the
19
+ weekly retrain pipeline in [remote-agent-dev-platform](https://github.com/Monibee-Fudgekins/remote-agent-dev-platform).
20
 
21
  - **Developed by:** Monibee-Fudgekins
22
+ - **License:** gemma (inherited from the base model)
23
+ - **Finetuned from model:** google/gemma-4-26B-A4B-it
24
 
25
  This gemma4 model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
26
 
27
+ > Promotions to this repo pass an eval gate: tuned pass@1 must meet the target and not
28
+ > regress vs the frozen base on a sandboxed multi-language coding suite. The full model
29
+ > card (eval table, training procedure) is rewritten by the pipeline on each promotion.
30
+
31
  [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)