OpenTransformer commited on
Commit
bf3c9a0
·
verified ·
1 Parent(s): 54b44d2

Clarify canonical SFT checkpoint and failed v3 archive

Browse files
Files changed (1) hide show
  1. README.md +6 -0
README.md CHANGED
@@ -14,6 +14,12 @@ Continuation of [AGILLM-3-Large](https://huggingface.co/OpenTransformer/AGILLM-3
14
 
15
  **Demo Space:** [OpenTransformer/AGILLM-3-large-v2-demo](https://huggingface.co/spaces/OpenTransformer/AGILLM-3-large-v2-demo)
16
 
 
 
 
 
 
 
17
  ## What happened to v1?
18
 
19
  A `transformers` library update (to v5.3.0) on 2026-03-11 silently broke the DeepSeek-V3.2 tokenizer's encode/decode pipeline:
 
14
 
15
  **Demo Space:** [OpenTransformer/AGILLM-3-large-v2-demo](https://huggingface.co/spaces/OpenTransformer/AGILLM-3-large-v2-demo)
16
 
17
+ ## Current recommended checkpoint (2026-05-21)
18
+
19
+ **Use [sft_chat_1024_220k_20260520/final.pt](sft_chat_1024_220k_20260520/final.pt) in AR mode.** This is the current canonical AGILLM-3 v2 chat checkpoint: it answers 1+1=2, 2+2=4, and 47+28=75 in the smoke rerun.
20
+
21
+ **Do not use [sft_sat_repair_v3_20260521/final.pt](sft_sat_repair_v3_20260521/final.pt) except as a failed-experiment archive.** That run regressed AR arithmetic and did not repair SAT. SAT mode remains experimental/broken pending a separate objective/inference fix.
22
+
23
  ## What happened to v1?
24
 
25
  A `transformers` library update (to v5.3.0) on 2026-03-11 silently broke the DeepSeek-V3.2 tokenizer's encode/decode pipeline: