lm-deluge 0.0.26__py3-none-any.whl → 0.0.28__py3-none-any.whl

lm_deluge/api_requests/gemini.py

@@ -37,14 +37,17 @@ async def _build_gemini_request(
 
     # Handle reasoning models (thinking)
     if model.reasoning_model:
-        request_json["generationConfig"]["thinkingConfig"] = {"includeThoughts": True}
-        if sampling_params.reasoning_effort and "flash" in model.id:
-            budget = {"low": 1024, "medium": 4096, "high": 16384}.get(
-                sampling_params.reasoning_effort
-            )
-            request_json["generationConfig"]["thinkingConfig"]["thinkingBudget"] = (
-                budget
-            )
+        thinking_config = None
+        effort = sampling_params.reasoning_effort
+        if effort is None or effort == "none":
+            # Explicitly disable thoughts when no effort is requested
+            thinking_config = {"includeThoughts": False, "thinkingBudget": 0}
+        else:
+            thinking_config = {"includeThoughts": True}
+            if effort in {"low", "medium", "high"} and "flash" in model.id:
+                budget = {"low": 1024, "medium": 4096, "high": 16384}[effort]
+                thinking_config["thinkingBudget"] = budget
+        request_json["generationConfig"]["thinkingConfig"] = thinking_config
 
     else:
         if sampling_params.reasoning_effort:

lm_deluge/api_requests/openai.py

@@ -37,7 +37,14 @@ async def _build_oa_chat_request(
     if model.reasoning_model:
         request_json["temperature"] = 1.0
         request_json["top_p"] = 1.0
-        request_json["reasoning_effort"] = sampling_params.reasoning_effort
+        effort = sampling_params.reasoning_effort
+        if effort in [None, "none"]:
+            # Disable reasoning for Gemini models when no effort requested
+            if "gemini" in model.id:
+                effort = "none"
+            else:
+                effort = "low"
+        request_json["reasoning_effort"] = effort
     else:
         if sampling_params.reasoning_effort:
             warnings.warn(

{lm_deluge-0.0.26.dist-info → lm_deluge-0.0.28.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lm_deluge
-Version: 0.0.26
+Version: 0.0.28
 Summary: Python utility for using LLM API models.
 Author-email: Benjamin Anderson <ben@trytaylor.ai>
 Requires-Python: >=3.10
@@ -274,7 +274,7 @@ We support all models in `src/lm_deluge/models.py`. Vertex support is not planne
 
 ## Feature Support
 
-We support structured outputs via `json_mode` parameter provided to `SamplingParams`. Structured outputs with a schema are planned. Reasoning models are supported via the `reasoning_effort` parameter, which is translated to a thinking budget for Claude/Gemini. Image models are supported. We support tool use as documented above. We support logprobs for OpenAI models that return them.
+We support structured outputs via `json_mode` parameter provided to `SamplingParams`. Structured outputs with a schema are planned. Reasoning models are supported via the `reasoning_effort` parameter, which is translated to a thinking budget for Claude/Gemini. Passing `None` (or the string `"none"`) disables Gemini thoughts entirely. Image models are supported. We support tool use as documented above. We support logprobs for OpenAI models that return them.
 
 ## Built‑in tools
 

{lm_deluge-0.0.26.dist-info → lm_deluge-0.0.28.dist-info}/RECORD

@@ -21,9 +21,9 @@ lm_deluge/api_requests/anthropic.py,sha256=Vi2iGQw6LGPW8d4X489Jg-dlEye529jf8i5Ys
 lm_deluge/api_requests/base.py,sha256=EVHNFtlttKbN7Tt1MnLaO-NjvKHPSV5CqlRv-OnpVAE,5593
 lm_deluge/api_requests/bedrock.py,sha256=p6leW5stnvb406lwbmVFmfTGxdEI-t0GfaTExpol3qk,10900
 lm_deluge/api_requests/common.py,sha256=BZ3vRO5TB669_UsNKugkkuFSzoLHOYJIKt4nV4sf4vc,422
-lm_deluge/api_requests/gemini.py,sha256=UeA4gsIePsIu6IQQ7izR359LLQQTi0Qky_ykjH1XQf4,7461
+lm_deluge/api_requests/gemini.py,sha256=W4NjQ0buBsdS7RYpzDahrXNQWMzDHRMLNRSphCOmIqg,7685
 lm_deluge/api_requests/mistral.py,sha256=S_LpOfCGbCVEROH_od3P-tYeNYTKFMamMTL-c_wFCBI,4597
-lm_deluge/api_requests/openai.py,sha256=jb_pBGqSbqs7SvbA45Odu3JkwS3jsKn-p3hG8-qkYbc,21509
+lm_deluge/api_requests/openai.py,sha256=hsJIMRO4wpalrczD0bVc--RWFu2BoXEp0USAwRlLQEA,21763
 lm_deluge/api_requests/response.py,sha256=FtkVYk_rDH93Kj9pqbB-l7a4dQHzVr6ivKL9khYKLbs,5966
 lm_deluge/api_requests/deprecated/bedrock.py,sha256=WrcIShCoO8JCUSlFOCHxg6KQCNTZfw3TpYTvSpYk4mA,11320
 lm_deluge/api_requests/deprecated/cohere.py,sha256=KgDScD6_bWhAzOY5BHZQKSA3kurt4KGENqC4wLsGmcU,5142
@@ -48,8 +48,8 @@ lm_deluge/util/logprobs.py,sha256=UkBZakOxWluaLqHrjARu7xnJ0uCHVfLGHJdnYlEcutk,11
 lm_deluge/util/spatial.py,sha256=BsF_UKhE-x0xBirc-bV1xSKZRTUhsOBdGqsMKme20C8,4099
 lm_deluge/util/validation.py,sha256=hz5dDb3ebvZrZhnaWxOxbNSVMI6nmaOODBkk0htAUhs,1575
 lm_deluge/util/xml.py,sha256=Ft4zajoYBJR3HHCt2oHwGfymGLdvp_gegVmJ-Wqk4Ck,10547
-lm_deluge-0.0.26.dist-info/licenses/LICENSE,sha256=uNNXGXPCw2TC7CUs7SEBkA-Mz6QBQFWUUEWDMgEs1dU,1058
-lm_deluge-0.0.26.dist-info/METADATA,sha256=2zzPvIBE0SdA0RAAKGnyUQDIq-sE_Eo34pfRu_ygi4I,12978
-lm_deluge-0.0.26.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-lm_deluge-0.0.26.dist-info/top_level.txt,sha256=hqU-TJX93yBwpgkDtYcXyLr3t7TLSCCZ_reytJjwBaE,10
-lm_deluge-0.0.26.dist-info/RECORD,,
+lm_deluge-0.0.28.dist-info/licenses/LICENSE,sha256=uNNXGXPCw2TC7CUs7SEBkA-Mz6QBQFWUUEWDMgEs1dU,1058
+lm_deluge-0.0.28.dist-info/METADATA,sha256=1yT_gHlhOwcS8hEC2kFnmxEN6jr9RBozCuaivHvO8vQ,13053
+lm_deluge-0.0.28.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+lm_deluge-0.0.28.dist-info/top_level.txt,sha256=hqU-TJX93yBwpgkDtYcXyLr3t7TLSCCZ_reytJjwBaE,10
+lm_deluge-0.0.28.dist-info/RECORD,,