openlit 1.16.0__py3-none-any.whl → 1.17.0__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (26) hide show
  1. openlit/instrumentation/anthropic/anthropic.py +30 -12
  2. openlit/instrumentation/anthropic/async_anthropic.py +29 -12
  3. openlit/instrumentation/bedrock/__init__.py +3 -3
  4. openlit/instrumentation/bedrock/bedrock.py +77 -307
  5. openlit/instrumentation/cohere/cohere.py +35 -14
  6. openlit/instrumentation/elevenlabs/async_elevenlabs.py +6 -2
  7. openlit/instrumentation/elevenlabs/elevenlabs.py +6 -2
  8. openlit/instrumentation/gpt4all/gpt4all.py +30 -10
  9. openlit/instrumentation/groq/async_groq.py +41 -21
  10. openlit/instrumentation/groq/groq.py +41 -21
  11. openlit/instrumentation/mistral/async_mistral.py +37 -16
  12. openlit/instrumentation/mistral/mistral.py +37 -16
  13. openlit/instrumentation/ollama/async_ollama.py +57 -20
  14. openlit/instrumentation/ollama/ollama.py +57 -20
  15. openlit/instrumentation/openai/async_azure_openai.py +106 -47
  16. openlit/instrumentation/openai/async_openai.py +78 -37
  17. openlit/instrumentation/openai/azure_openai.py +101 -43
  18. openlit/instrumentation/openai/openai.py +78 -39
  19. openlit/instrumentation/transformers/transformers.py +21 -17
  20. openlit/instrumentation/vertexai/async_vertexai.py +104 -35
  21. openlit/instrumentation/vertexai/vertexai.py +104 -35
  22. openlit/semcov/__init__.py +6 -4
  23. {openlit-1.16.0.dist-info → openlit-1.17.0.dist-info}/METADATA +2 -2
  24. {openlit-1.16.0.dist-info → openlit-1.17.0.dist-info}/RECORD +26 -26
  25. {openlit-1.16.0.dist-info → openlit-1.17.0.dist-info}/LICENSE +0 -0
  26. {openlit-1.16.0.dist-info → openlit-1.17.0.dist-info}/WHEEL +0 -0
openlit/instrumentation/cohere/cohere.py CHANGED
@@ -89,8 +89,12 @@ def embed(gen_ai_endpoint, version, environment, application_name, tracer,
89
89
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
90
90
  cost)
91
91
  if trace_content:
92
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
93
- prompt)
92
+ span.add_event(
93
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
94
+ attributes={
95
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
96
+ },
97
+ )
94
98
 
95
99
  span.set_status(Status(StatusCode.OK))
96
100
 
@@ -193,7 +197,7 @@ def chat(gen_ai_endpoint, version, environment, application_name, tracer,
193
197
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
194
198
  kwargs.get("temperature", 0.3))
195
199
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
196
- kwargs.get("max_tokens", ""))
200
+ kwargs.get("max_tokens", -1))
197
201
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
198
202
  kwargs.get("seed", ""))
199
203
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
@@ -205,7 +209,7 @@ def chat(gen_ai_endpoint, version, environment, application_name, tracer,
205
209
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
206
210
  response.generation_id)
207
211
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
208
- response.finish_reason)
212
+ [response.finish_reason])
209
213
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
210
214
  response.meta.billed_units.input_tokens)
211
215
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
@@ -215,11 +219,20 @@ def chat(gen_ai_endpoint, version, environment, application_name, tracer,
215
219
  response.meta.billed_units.output_tokens)
216
220
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
217
221
  cost)
222
+
218
223
  if trace_content:
219
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
220
- kwargs.get("message", ""))
221
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
222
- response.text)
224
+ span.add_event(
225
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
226
+ attributes={
227
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: kwargs.get("message", ""),
228
+ },
229
+ )
230
+ span.add_event(
231
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
232
+ attributes={
233
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.text,
234
+ },
235
+ )
223
236
 
224
237
  span.set_status(Status(StatusCode.OK))
225
238
 
@@ -336,7 +349,7 @@ def chat_stream(gen_ai_endpoint, version, environment, application_name,
336
349
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
337
350
  kwargs.get("temperature", 0.3))
338
351
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
339
- kwargs.get("max_tokens", ""))
352
+ kwargs.get("max_tokens", -1))
340
353
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
341
354
  kwargs.get("seed", ""))
342
355
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
@@ -348,7 +361,7 @@ def chat_stream(gen_ai_endpoint, version, environment, application_name,
348
361
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_ID,
349
362
  response_id)
350
363
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
351
- finish_reason)
364
+ [finish_reason])
352
365
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
353
366
  prompt_tokens)
354
367
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
@@ -358,10 +371,18 @@ def chat_stream(gen_ai_endpoint, version, environment, application_name,
358
371
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
359
372
  cost)
360
373
  if trace_content:
361
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
362
- kwargs.get("message", ""))
363
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
364
- llmresponse)
374
+ span.add_event(
375
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
376
+ attributes={
377
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: kwargs.get("message", ""),
378
+ },
379
+ )
380
+ span.add_event(
381
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
382
+ attributes={
383
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
384
+ },
385
+ )
365
386
 
366
387
  span.set_status(Status(StatusCode.OK))
367
388
 
openlit/instrumentation/elevenlabs/async_elevenlabs.py CHANGED
@@ -80,8 +80,12 @@ def async_generate(gen_ai_endpoint, version, environment, application_name,
80
80
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
81
81
  cost)
82
82
  if trace_content:
83
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
84
- str(kwargs.get("text", "")))
83
+ span.add_event(
84
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
85
+ attributes={
86
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: str(kwargs.get("text", "")),
87
+ },
88
+ )
85
89
 
86
90
  span.set_status(Status(StatusCode.OK))
87
91
 
openlit/instrumentation/elevenlabs/elevenlabs.py CHANGED
@@ -86,8 +86,12 @@ def generate(gen_ai_endpoint, version, environment, application_name,
86
86
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
87
87
  cost)
88
88
  if trace_content:
89
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
90
- str(kwargs.get("text", "")))
89
+ span.add_event(
90
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
91
+ attributes={
92
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: str(kwargs.get("text", "")),
93
+ },
94
+ )
91
95
 
92
96
  span.set_status(Status(StatusCode.OK))
93
97
 
openlit/instrumentation/gpt4all/gpt4all.py CHANGED
@@ -107,10 +107,18 @@ def generate(gen_ai_endpoint, version, environment, application_name,
107
107
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
108
108
  True)
109
109
  if trace_content:
110
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
111
- prompt)
112
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
113
- llmresponse)
110
+ span.add_event(
111
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
112
+ attributes={
113
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
114
+ },
115
+ )
116
+ span.add_event(
117
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
118
+ attributes={
119
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
120
+ },
121
+ )
114
122
 
115
123
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
116
124
  prompt_tokens)
@@ -195,10 +203,18 @@ def generate(gen_ai_endpoint, version, environment, application_name,
195
203
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
196
204
  False)
197
205
  if trace_content:
198
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
199
- prompt)
200
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
201
- response)
206
+ span.add_event(
207
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
208
+ attributes={
209
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
210
+ },
211
+ )
212
+ span.add_event(
213
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
214
+ attributes={
215
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response,
216
+ },
217
+ )
202
218
 
203
219
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
204
220
  prompt_tokens)
@@ -313,8 +329,12 @@ def embed(gen_ai_endpoint, version, environment, application_name,
313
329
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
314
330
  cost)
315
331
  if trace_content:
316
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
317
- prompt)
332
+ span.add_event(
333
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
334
+ attributes={
335
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
336
+ },
337
+ )
318
338
 
319
339
  span.set_status(Status(StatusCode.OK))
320
340
 
openlit/instrumentation/groq/async_groq.py CHANGED
@@ -119,15 +119,15 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
119
119
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_USER,
120
120
  kwargs.get("user", ""))
121
121
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
122
- kwargs.get("top_p", 1))
122
+ kwargs.get("top_p", 1.0))
123
123
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
124
- kwargs.get("max_tokens", ""))
124
+ kwargs.get("max_tokens", -1))
125
125
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
126
- kwargs.get("temperature", 1))
126
+ kwargs.get("temperature", 1.0))
127
127
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
128
- kwargs.get("presence_penalty", 0))
128
+ kwargs.get("presence_penalty", 0.0))
129
129
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
130
- kwargs.get("frequency_penalty", 0))
130
+ kwargs.get("frequency_penalty", 0.0))
131
131
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
132
132
  kwargs.get("seed", ""))
133
133
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
@@ -141,10 +141,18 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
141
141
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
142
142
  cost)
143
143
  if trace_content:
144
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
145
- prompt)
146
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
147
- llmresponse)
144
+ span.add_event(
145
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
146
+ attributes={
147
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
148
+ },
149
+ )
150
+ span.add_event(
151
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
152
+ attributes={
153
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
154
+ },
155
+ )
148
156
 
149
157
  span.set_status(Status(StatusCode.OK))
150
158
 
@@ -221,24 +229,28 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
221
229
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
222
230
  kwargs.get("model", "llama3-8b-8192"))
223
231
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
224
- kwargs.get("top_p", 1))
232
+ kwargs.get("top_p", 1.0))
225
233
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
226
- kwargs.get("max_tokens", ""))
234
+ kwargs.get("max_tokens", -1))
227
235
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_USER,
228
236
  kwargs.get("name", ""))
229
237
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
230
- kwargs.get("temperature", 1))
238
+ kwargs.get("temperature", 1.0))
231
239
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
232
- kwargs.get("presence_penalty", 0))
240
+ kwargs.get("presence_penalty", 0.0))
233
241
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
234
- kwargs.get("frequency_penalty", 0))
242
+ kwargs.get("frequency_penalty", 0.0))
235
243
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
236
244
  kwargs.get("seed", ""))
237
245
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
238
246
  False)
239
247
  if trace_content:
240
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
241
- prompt)
248
+ span.add_event(
249
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
250
+ attributes={
251
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
252
+ },
253
+ )
242
254
 
243
255
  # Set span attributes when tools is not passed to the function call
244
256
  if "tools" not in kwargs:
@@ -254,23 +266,31 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
254
266
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
255
267
  response.usage.total_tokens)
256
268
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
257
- response.choices[0].finish_reason)
269
+ [response.choices[0].finish_reason])
258
270
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
259
271
  cost)
260
272
 
261
273
  # Set span attributes for when n = 1 (default)
262
274
  if "n" not in kwargs or kwargs["n"] == 1:
263
275
  if trace_content:
264
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
265
- response.choices[0].message.content)
276
+ span.add_event(
277
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
278
+ attributes={
279
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[0].message.content,
280
+ },
281
+ )
266
282
 
267
283
  # Set span attributes for when n > 0
268
284
  else:
269
285
  i = 0
270
286
  while i < kwargs["n"] and trace_content is True:
271
287
  attribute_name = f"gen_ai.completion.{i}"
272
- span.set_attribute(attribute_name,
273
- response.choices[i].message.content)
288
+ span.add_event(
289
+ name=attribute_name,
290
+ attributes={
291
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[i].message.content,
292
+ },
293
+ )
274
294
  i += 1
275
295
 
276
296
  # Return original response
openlit/instrumentation/groq/groq.py CHANGED
@@ -119,15 +119,15 @@ def chat(gen_ai_endpoint, version, environment, application_name,
119
119
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_USER,
120
120
  kwargs.get("user", ""))
121
121
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
122
- kwargs.get("top_p", 1))
122
+ kwargs.get("top_p", 1.0))
123
123
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
124
- kwargs.get("max_tokens", ""))
124
+ kwargs.get("max_tokens", -1))
125
125
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
126
- kwargs.get("temperature", 1))
126
+ kwargs.get("temperature", 1.0))
127
127
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
128
- kwargs.get("presence_penalty", 0))
128
+ kwargs.get("presence_penalty", 0.0))
129
129
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
130
- kwargs.get("frequency_penalty", 0))
130
+ kwargs.get("frequency_penalty", 0.0))
131
131
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
132
132
  kwargs.get("seed", ""))
133
133
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
@@ -141,10 +141,18 @@ def chat(gen_ai_endpoint, version, environment, application_name,
141
141
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
142
142
  cost)
143
143
  if trace_content:
144
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
145
- prompt)
146
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
147
- llmresponse)
144
+ span.add_event(
145
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
146
+ attributes={
147
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
148
+ },
149
+ )
150
+ span.add_event(
151
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
152
+ attributes={
153
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
154
+ },
155
+ )
148
156
 
149
157
  span.set_status(Status(StatusCode.OK))
150
158
 
@@ -221,24 +229,28 @@ def chat(gen_ai_endpoint, version, environment, application_name,
221
229
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MODEL,
222
230
  kwargs.get("model", "llama3-8b-8192"))
223
231
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
224
- kwargs.get("top_p", 1))
232
+ kwargs.get("top_p", 1.0))
225
233
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
226
- kwargs.get("max_tokens", ""))
234
+ kwargs.get("max_tokens", -1))
227
235
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_USER,
228
236
  kwargs.get("name", ""))
229
237
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
230
- kwargs.get("temperature", 1))
238
+ kwargs.get("temperature", 1.0))
231
239
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_PRESENCE_PENALTY,
232
- kwargs.get("presence_penalty", 0))
240
+ kwargs.get("presence_penalty", 0.0))
233
241
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_FREQUENCY_PENALTY,
234
- kwargs.get("frequency_penalty", 0))
242
+ kwargs.get("frequency_penalty", 0.0))
235
243
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
236
244
  kwargs.get("seed", ""))
237
245
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
238
246
  False)
239
247
  if trace_content:
240
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
241
- prompt)
248
+ span.add_event(
249
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
250
+ attributes={
251
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
252
+ },
253
+ )
242
254
 
243
255
  # Set span attributes when tools is not passed to the function call
244
256
  if "tools" not in kwargs:
@@ -254,23 +266,31 @@ def chat(gen_ai_endpoint, version, environment, application_name,
254
266
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_TOTAL_TOKENS,
255
267
  response.usage.total_tokens)
256
268
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
257
- response.choices[0].finish_reason)
269
+ [response.choices[0].finish_reason])
258
270
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
259
271
  cost)
260
272
 
261
273
  # Set span attributes for when n = 1 (default)
262
274
  if "n" not in kwargs or kwargs["n"] == 1:
263
275
  if trace_content:
264
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
265
- response.choices[0].message.content)
276
+ span.add_event(
277
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
278
+ attributes={
279
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[0].message.content,
280
+ },
281
+ )
266
282
 
267
283
  # Set span attributes for when n > 0
268
284
  else:
269
285
  i = 0
270
286
  while i < kwargs["n"] and trace_content is True:
271
287
  attribute_name = f"gen_ai.completion.{i}"
272
- span.set_attribute(attribute_name,
273
- response.choices[i].message.content)
288
+ span.add_event(
289
+ name=attribute_name,
290
+ attributes={
291
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[i].message.content,
292
+ },
293
+ )
274
294
  i += 1
275
295
 
276
296
  # Return original response
openlit/instrumentation/mistral/async_mistral.py CHANGED
@@ -95,15 +95,15 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
95
95
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
96
96
  kwargs.get("temperature", 0.7))
97
97
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
98
- kwargs.get("top_p", 1))
98
+ kwargs.get("top_p", 1.0))
99
99
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
100
- kwargs.get("max_tokens", ""))
100
+ kwargs.get("max_tokens", -1))
101
101
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
102
102
  kwargs.get("random_seed", ""))
103
103
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
104
104
  False)
105
105
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
106
- response.choices[0].finish_reason)
106
+ [response.choices[0].finish_reason])
107
107
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
108
108
  response.usage.prompt_tokens)
109
109
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
@@ -113,10 +113,19 @@ def async_chat(gen_ai_endpoint, version, environment, application_name,
113
113
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
114
114
  cost)
115
115
  if trace_content:
116
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
117
- prompt)
118
- # pylint: disable=line-too-long
119
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION, response.choices[0].message.content if response.choices[0].message.content else "")
116
+ span.add_event(
117
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
118
+ attributes={
119
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
120
+ },
121
+ )
122
+ span.add_event(
123
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
124
+ attributes={
125
+ # pylint: disable=line-too-long
126
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[0].message.content if response.choices[0].message.content else "",
127
+ },
128
+ )
120
129
 
121
130
  span.set_status(Status(StatusCode.OK))
122
131
 
@@ -252,15 +261,15 @@ def async_chat_stream(gen_ai_endpoint, version, environment, application_name,
252
261
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
253
262
  kwargs.get("temperature", 0.7))
254
263
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
255
- kwargs.get("top_p", 1))
264
+ kwargs.get("top_p", 1.0))
256
265
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
257
- kwargs.get("max_tokens", ""))
266
+ kwargs.get("max_tokens", -1))
258
267
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
259
268
  kwargs.get("random_seed", ""))
260
269
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
261
270
  True)
262
271
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
263
- finish_reason)
272
+ [finish_reason])
264
273
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
265
274
  prompt_tokens)
266
275
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
@@ -270,10 +279,18 @@ def async_chat_stream(gen_ai_endpoint, version, environment, application_name,
270
279
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
271
280
  cost)
272
281
  if trace_content:
273
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
274
- prompt)
275
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
276
- llmresponse)
282
+ span.add_event(
283
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
284
+ attributes={
285
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
286
+ },
287
+ )
288
+ span.add_event(
289
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
290
+ attributes={
291
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
292
+ },
293
+ )
277
294
 
278
295
  span.set_status(Status(StatusCode.OK))
279
296
 
@@ -378,8 +395,12 @@ def async_embeddings(gen_ai_endpoint, version, environment, application_name,
378
395
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
379
396
  cost)
380
397
  if trace_content:
381
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
382
- prompt)
398
+ span.add_event(
399
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
400
+ attributes={
401
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
402
+ },
403
+ )
383
404
 
384
405
  span.set_status(Status(StatusCode.OK))
385
406
 
openlit/instrumentation/mistral/mistral.py CHANGED
@@ -94,15 +94,15 @@ def chat(gen_ai_endpoint, version, environment, application_name,
94
94
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
95
95
  kwargs.get("temperature", 0.7))
96
96
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
97
- kwargs.get("top_p", 1))
97
+ kwargs.get("top_p", 1.0))
98
98
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
99
- kwargs.get("max_tokens", ""))
99
+ kwargs.get("max_tokens", -1))
100
100
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
101
101
  kwargs.get("random_seed", ""))
102
102
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
103
103
  False)
104
104
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
105
- response.choices[0].finish_reason)
105
+ [response.choices[0].finish_reason])
106
106
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
107
107
  response.usage.prompt_tokens)
108
108
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
@@ -112,10 +112,19 @@ def chat(gen_ai_endpoint, version, environment, application_name,
112
112
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
113
113
  cost)
114
114
  if trace_content:
115
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
116
- prompt)
117
- # pylint: disable=line-too-long
118
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION, response.choices[0].message.content if response.choices[0].message.content else "")
115
+ span.add_event(
116
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
117
+ attributes={
118
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
119
+ },
120
+ )
121
+ span.add_event(
122
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
123
+ # pylint: disable=line-too-long
124
+ attributes={
125
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: response.choices[0].message.content if response.choices[0].message.content else "",
126
+ },
127
+ )
119
128
 
120
129
  span.set_status(Status(StatusCode.OK))
121
130
 
@@ -251,15 +260,15 @@ def chat_stream(gen_ai_endpoint, version, environment, application_name,
251
260
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TEMPERATURE,
252
261
  kwargs.get("temperature", 0.7))
253
262
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_TOP_P,
254
- kwargs.get("top_p", 1))
263
+ kwargs.get("top_p", 1.0))
255
264
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_MAX_TOKENS,
256
- kwargs.get("max_tokens", ""))
265
+ kwargs.get("max_tokens", -1))
257
266
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_SEED,
258
267
  kwargs.get("random_seed", ""))
259
268
  span.set_attribute(SemanticConvetion.GEN_AI_REQUEST_IS_STREAM,
260
269
  True)
261
270
  span.set_attribute(SemanticConvetion.GEN_AI_RESPONSE_FINISH_REASON,
262
- finish_reason)
271
+ [finish_reason])
263
272
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_PROMPT_TOKENS,
264
273
  prompt_tokens)
265
274
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COMPLETION_TOKENS,
@@ -269,10 +278,18 @@ def chat_stream(gen_ai_endpoint, version, environment, application_name,
269
278
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
270
279
  cost)
271
280
  if trace_content:
272
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
273
- prompt)
274
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_COMPLETION,
275
- llmresponse)
281
+ span.add_event(
282
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
283
+ attributes={
284
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
285
+ },
286
+ )
287
+ span.add_event(
288
+ name=SemanticConvetion.GEN_AI_CONTENT_COMPLETION_EVENT,
289
+ attributes={
290
+ SemanticConvetion.GEN_AI_CONTENT_COMPLETION: llmresponse,
291
+ },
292
+ )
276
293
 
277
294
  span.set_status(Status(StatusCode.OK))
278
295
 
@@ -377,8 +394,12 @@ def embeddings(gen_ai_endpoint, version, environment, application_name,
377
394
  span.set_attribute(SemanticConvetion.GEN_AI_USAGE_COST,
378
395
  cost)
379
396
  if trace_content:
380
- span.set_attribute(SemanticConvetion.GEN_AI_CONTENT_PROMPT,
381
- prompt)
397
+ span.add_event(
398
+ name=SemanticConvetion.GEN_AI_CONTENT_PROMPT_EVENT,
399
+ attributes={
400
+ SemanticConvetion.GEN_AI_CONTENT_PROMPT: prompt,
401
+ },
402
+ )
382
403
 
383
404
  span.set_status(Status(StatusCode.OK))
384
405