Add a few items to faq (#2734)

* Add a few items to faq

* formatting

* chore: lint

Author: winglian

docs/faq.qmd

@@ -110,3 +110,17 @@ description: Frequently asked questions
 > A: If `eot_tokens: ` is not provided, the default behavior is the same as before. EOS tokens used to delimit turns are masked/unmasked depending on whether the turn is trainable.
 
 > Internally, `eot_tokens: tokenizer.eos_token` and `train_on_eot: train_on_eos` (which defaults to `turn`). This transition helps clarify the naming and behavior of EOT/EOS tokens.
+
+**Q: `Data processing error: CAS service error`**
+
+> A: Try disabling XET with `export HF_HUB_DISABLE_XET=1`
+
+**Q: `torch._inductor.exc.LoweringException: NoValidChoicesError: No choices to select, please consider adding ATEN into max_autotune_gemm_backends config (defined in torch/_inductor/config.py) to allow at least one choice. `**
+
+> A: Depending on the version of torch, you may need to include this in your YAML:
+
+> ```yaml
+> flex_attn_compile_kwargs:
+>   dynamic: false
+>   mode: max-autotune-no-cudagraphs
+> ```

src/axolotl/utils/schemas/config.py

@@ -1175,18 +1175,6 @@ def check_kto_config(cls, data):
 
         return data
 
-    # @model_validator(mode="before")
-    # @classmethod
-    # def check_grpo_peft_liger(cls, data):
-    #     if (
-    #         data.get("rl") == "grpo"
-    #         and data.get("trl", {})
-    #         and data.get("trl").get("use_liger_loss")
-    #         and data.get("adapter")
-    #     ):
-    #         raise ValueError("PEFT + GRPO + Liger is not yet supported")
-    #     return data
-    #
     @model_validator(mode="before")
     @classmethod
     def check_grpo_liger_sequence_parallel(cls, data):