added default inference config from unsloth notebooks

studio/backend/assets/configs/model_defaults/ernie/unsloth_ERNIE-4.5-VL-28B-A3B-PT.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/ERNIE-4.5-VL-28B-A3B-PT
 # Based on ERNIE_4_5_VL_28B_A3B_PT_Vision.ipynb
 # Also applies to: unsloth/ERNIE-4.5-VL-28B-A3B-PT
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -45,3 +46,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+

studio/backend/assets/configs/model_defaults/llama/unsloth_Llama-3.2-11B-Vision-Instruct.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/Llama-3.2-11B-Vision-Instruct
 # Based on Llama3.2_(11B)-Vision.ipynb
 # Also applies to: unsloth/Llama-3.2-11B-Vision-Instruct-unsloth-bnb-4bit, meta-llama/Llama-3.2-11B-Vision-Instruct, unsloth/Llama-3.2-11B-Vision-Instruct-bnb-4bit
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -39,3 +40,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+

studio/backend/assets/configs/model_defaults/llama/unsloth_Llama-3.2-3B-Instruct.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/Llama-3.2-3B-Instruct
 # Based on Llama3.2_(1B_and_3B)-Conversational.ipynb
 # Also applies to: unsloth/Llama-3.2-3B-Instruct-unsloth-bnb-4bit, meta-llama/Llama-3.2-3B-Instruct, unsloth/Llama-3.2-3B-Instruct-bnb-4bit, RedHatAI/Llama-3.2-3B-Instruct-FP8, unsloth/Llama-3.2-3B-Instruct-FP8-Block, unsloth/Llama-3.2-3B-Instruct-FP8-Dynamic
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -45,3 +46,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+

studio/backend/assets/configs/model_defaults/llama/unsloth_Llama-3.3-70B-Instruct.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/Llama-3.3-70B-Instruct
 # Based on Llama3.3_(70B)_A100-Conversational.ipynb
 # Also applies to: unsloth/Llama-3.3-70B-Instruct-unsloth-bnb-4bit, meta-llama/Llama-3.3-70B-Instruct, unsloth/Llama-3.3-70B-Instruct-bnb-4bit, RedHatAI/Llama-3.3-70B-Instruct-FP8, unsloth/Llama-3.3-70B-Instruct-FP8-Block, unsloth/Llama-3.3-70B-Instruct-FP8-Dynamic
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -45,3 +46,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+

studio/backend/assets/configs/model_defaults/llasa/unsloth_Llasa-3B.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/Llasa-3B
 # Based on Llasa_TTS_(3B).ipynb and Llasa_TTS_(1B).ipynb
 # Also applies to: HKUSTAudio/Llasa-1B
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -40,3 +41,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.2
+  top_p: 1.2
+

studio/backend/assets/configs/model_defaults/mistral/unsloth_Pixtral-12B-2409.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/Pixtral-12B-2409
 # Based on Pixtral_(12B)-Vision.ipynb
 # Also applies to: unsloth/Pixtral-12B-2409-unsloth-bnb-4bit, mistralai/Pixtral-12B-2409, unsloth/Pixtral-12B-2409-bnb-4bit
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -39,3 +40,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+

studio/backend/assets/configs/model_defaults/other/OuteAI_Llama-OuteTTS-1.0-1B.yaml

@@ -1,6 +1,7 @@
 # Model defaults for OuteAI/Llama-OuteTTS-1.0-1B
 # Based on Oute_TTS_(1B).ipynb
 # Also applies to: OuteAI/Llama-OuteTTS-1.0-1B
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -40,3 +41,9 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 0.4
+  top_k: 40
+  top_p: 0.9
+  min_p: 0.05
+

studio/backend/assets/configs/model_defaults/other/Spark-TTS-0.5B_LLM.yaml

@@ -1,6 +1,7 @@
 # Model defaults for Spark-TTS-0.5B/LLM
 # Based on Spark_TTS_(0_5B).ipynb
 # Also applies to: Spark-TTS-0.5B/LLM
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -45,3 +46,8 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 0.8
+  top_k: 50
+  top_p: 1.0
+

studio/backend/assets/configs/model_defaults/other/unsloth_LFM2-1.2B.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/LFM2-1.2B
 # Based on Liquid_LFM2_(1.2B)-Conversational.ipynb
 # Also applies to: unsloth/LFM2-1.2B
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -39,3 +40,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 0.3
+  min_p: 0.15
+

studio/backend/assets/configs/model_defaults/other/unsloth_PaddleOCR-VL.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/PaddleOCR-VL
 # Based on Paddle_OCR_(1B)_Vision.ipynb
 # Also applies to: unsloth/PaddleOCR-VL
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -45,3 +46,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+

studio/backend/assets/configs/model_defaults/other/unsloth_orpheus-3b-0.1-ft.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/orpheus-3b-0.1-ft
 # Based on Orpheus_(3B)-TTS.ipynb
 # Also applies to: unsloth/orpheus-3b-0.1-ft-unsloth-bnb-4bit, canopylabs/orpheus-3b-0.1-ft, unsloth/orpheus-3b-0.1-ft-bnb-4bit
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -45,3 +46,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 0.6
+  top_p: 0.95
+

studio/backend/assets/configs/model_defaults/qwen/unsloth_Qwen2-VL-7B-Instruct.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/Qwen2-VL-7B-Instruct
 # Based on Qwen2_VL_(7B)-Vision.ipynb 
 # Also applies to: unsloth/Qwen2-VL-7B-Instruct-unsloth-bnb-4bit, Qwen/Qwen2-VL-7B-Instruct, unsloth/Qwen2-VL-7B-Instruct-bnb-4bit
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -39,3 +40,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+

studio/backend/assets/configs/model_defaults/qwen/unsloth_Qwen2.5-Coder-14B-Instruct.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/Qwen2.5-Coder-14B-Instruct
 # Based on Qwen2.5_Coder_(14B)-Conversational.ipynb
 # Also applies to: unsloth/Qwen2.5-Coder-14B-Instruct-bnb-4bit, Qwen/Qwen2.5-Coder-14B-Instruct
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -45,3 +46,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+

studio/backend/assets/configs/model_defaults/qwen/unsloth_Qwen2.5-VL-7B-Instruct-bnb-4bit.yaml

@@ -1,6 +1,7 @@
 # Model defaults for unsloth/Qwen2.5-VL-7B-Instruct-bnb-4bit
 # Based on Qwen2.5_VL_(7B)-Vision.ipynb
 # Also applies to: unsloth/Qwen2.5-VL-7B-Instruct, Qwen/Qwen2.5-VL-7B-Instruct, unsloth/Qwen2.5-VL-7B-Instruct-unsloth-bnb-4bit
+# added inference parameters from unsloth notebook
 
 training:
   max_seq_length: 2048
@@ -39,3 +40,7 @@ logging:
   tensorboard_dir: "runs"
   log_frequency: 10
 
+inference:
+  temperature: 1.5
+  min_p: 0.1
+