Upload llmdolphin.py

llmdolphin.py

@@ -27,10 +27,14 @@ llm_models = {
     "Nemo-12B-Marlin-v5-Q4_K_M.gguf": ["starble-dev/Nemo-12B-Marlin-v5-GGUF", MessagesFormatterType.CHATML],
     "Nemo-12B-Marlin-v7.Q4_K_M.gguf": ["mradermacher/Nemo-12B-Marlin-v7-GGUF", MessagesFormatterType.MISTRAL],
     "NemoDori-v0.2-Upscaled.1-14B.Q4_K_M.gguf": ["mradermacher/NemoDori-v0.2-Upscaled.1-14B-GGUF", MessagesFormatterType.MISTRAL],
+    "Fireball-Mistral-Nemo-Base-2407-sft-v2.2a.Q4_K_M.gguf": ["mradermacher/Fireball-Mistral-Nemo-Base-2407-sft-v2.2a-GGUF", MessagesFormatterType.MISTRAL],
     "Trinas_Nectar-8B-model_stock.i1-Q4_K_M.gguf": ["mradermacher/Trinas_Nectar-8B-model_stock-i1-GGUF", MessagesFormatterType.MISTRAL],
+    "storiecreative-q5_k_m.gguf": ["ClaudioItaly/StorieCreative-Q5_K_M-GGUF", MessagesFormatterType.MISTRAL],
     "L3.1-Sithamo-v0.4-8B.q5_k_m.gguf": ["kromquant/L3.1-Siithamo-v0.4-8B-GGUFs", MessagesFormatterType.MISTRAL],
     "Berry-Spark-7B-Fix.Q5_K_M.gguf": ["mradermacher/Berry-Spark-7B-Fix-GGUF", MessagesFormatterType.OPEN_CHAT],
+    "llama3.1-gutenberg-8B.Q4_K_S.gguf": ["mradermacher/llama3.1-gutenberg-8B-GGUF", MessagesFormatterType.LLAMA_3],
     "L3.1-Romes-Ninomos.Q4_K_S.gguf": ["mradermacher/L3.1-Romes-Ninomos-GGUF", MessagesFormatterType.LLAMA_3],
+    "nemo-12b-summarizer-de-v3.Q4_K_M.gguf": ["mradermacher/nemo-12b-summarizer-de-v3-GGUF", MessagesFormatterType.MISTRAL],
     "suzume-llama-3-8B-multilingual-orpo-borda-top25.Q5_K_M.gguf": ["darkshapes/suzume-llama-3-8B-multilingual-orpo-borda-top25-gguf", MessagesFormatterType.LLAMA_3],
     "Fireball-Mistral-Nemo-Base-2407-sft-v2.1.Q4_K_M.gguf": ["mradermacher/Fireball-Mistral-Nemo-Base-2407-sft-v2.1-GGUF", MessagesFormatterType.MISTRAL],
     "gemma-2-9B-it-advanced-v2.1-Q5_K_M.gguf": ["jsgreenawalt/gemma-2-9B-it-advanced-v2.1-GGUF", MessagesFormatterType.ALPACA],
@@ -235,6 +239,7 @@ llm_models = {
     "ChatWaifu_v1.1.Q5_K_M.gguf": ["mradermacher/ChatWaifu_v1.1-GGUF", MessagesFormatterType.MISTRAL],
     "Ninja-V2-7B_Q4_K_M.gguf": ["Local-Novel-LLM-project/Ninja-V2-7B-GGUF", MessagesFormatterType.VICUNA],
     "Yamase-12B.Q4_K_M.gguf": ["mradermacher/Yamase-12B-GGUF", MessagesFormatterType.MISTRAL],
+    "borea-phi-3.5-mini-instruct-common.Q5_K_M.gguf": ["keitokei1994/Borea-Phi-3.5-mini-Instruct-Common-GGUF", MessagesFormatterType.PHI_3],
     "Llama-3-Nymeria-ELYZA-8B.i1-Q4_K_M.gguf": ["mradermacher/Llama-3-Nymeria-ELYZA-8B-i1-GGUF", MessagesFormatterType.LLAMA_3],
     "suzume-llama-3-8B-japanese.Q4_K_M.gguf": ["PrunaAI/lightblue-suzume-llama-3-8B-japanese-GGUF-smashed", MessagesFormatterType.LLAMA_3],
     "suzume-llama-3-8B-multilingual-orpo-borda-top25.Q4_K_M.gguf": ["RichardErkhov/lightblue_-_suzume-llama-3-8B-multilingual-orpo-borda-top25-gguf", MessagesFormatterType.LLAMA_3],
@@ -247,6 +252,7 @@ llm_models = {
     #"": ["", MessagesFormatterType.ALPACA],
     #"": ["", MessagesFormatterType.OPEN_CHAT],
     #"": ["", MessagesFormatterType.CHATML],
+    #"": ["", MessagesFormatterType.PHI_3],
 }
 llm_formats = {
     "MISTRAL": MessagesFormatterType.MISTRAL,
@@ -649,7 +655,7 @@ def get_raw_prompt(msg: str):
     return re.sub(r'[*/:_"#\n]', ' ', ", ".join(m)).lower() if m else ""
 
 
-@spaces.GPU
+@spaces.GPU(duration=30)
 def dolphin_respond(
     message: str,
     history: list[tuple[str, str]],
@@ -742,7 +748,7 @@ def dolphin_parse(
     return ", ".join(prompts), gr.update(interactive=True), gr.update(interactive=True)
 
 
-@spaces.GPU
+@spaces.GPU(duration=30)
 def dolphin_respond_auto(
     message: str,
     history: list[tuple[str, str]],