chainlit 1.0.400__py3-none-any.whl → 2.0.3__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.

Potentially problematic release.

This version of chainlit might be problematic. Click here for more details.

Files changed (113) hide show
  1. chainlit/__init__.py +98 -279
  2. chainlit/_utils.py +8 -0
  3. chainlit/action.py +12 -10
  4. chainlit/{auth.py → auth/__init__.py} +28 -36
  5. chainlit/auth/cookie.py +122 -0
  6. chainlit/auth/jwt.py +39 -0
  7. chainlit/cache.py +4 -6
  8. chainlit/callbacks.py +362 -0
  9. chainlit/chat_context.py +64 -0
  10. chainlit/chat_settings.py +3 -1
  11. chainlit/cli/__init__.py +77 -8
  12. chainlit/config.py +181 -101
  13. chainlit/context.py +42 -13
  14. chainlit/copilot/dist/index.js +8750 -903
  15. chainlit/data/__init__.py +101 -416
  16. chainlit/data/acl.py +6 -2
  17. chainlit/data/base.py +107 -0
  18. chainlit/data/chainlit_data_layer.py +608 -0
  19. chainlit/data/dynamodb.py +590 -0
  20. chainlit/data/literalai.py +500 -0
  21. chainlit/data/sql_alchemy.py +721 -0
  22. chainlit/data/storage_clients/__init__.py +0 -0
  23. chainlit/data/storage_clients/azure.py +81 -0
  24. chainlit/data/storage_clients/azure_blob.py +89 -0
  25. chainlit/data/storage_clients/base.py +26 -0
  26. chainlit/data/storage_clients/gcs.py +88 -0
  27. chainlit/data/storage_clients/s3.py +75 -0
  28. chainlit/data/utils.py +29 -0
  29. chainlit/discord/__init__.py +6 -0
  30. chainlit/discord/app.py +354 -0
  31. chainlit/element.py +91 -33
  32. chainlit/emitter.py +80 -29
  33. chainlit/frontend/dist/assets/DailyMotion-C_XC7xJI.js +1 -0
  34. chainlit/frontend/dist/assets/Dataframe-Cs4l4hA1.js +22 -0
  35. chainlit/frontend/dist/assets/Facebook-CUeCH7hk.js +1 -0
  36. chainlit/frontend/dist/assets/FilePlayer-CB-fYkx8.js +1 -0
  37. chainlit/frontend/dist/assets/Kaltura-YX6qaq72.js +1 -0
  38. chainlit/frontend/dist/assets/Mixcloud-DGV0ldjP.js +1 -0
  39. chainlit/frontend/dist/assets/Mux-CmRss5oc.js +1 -0
  40. chainlit/frontend/dist/assets/Preview-DBVJn7-H.js +1 -0
  41. chainlit/frontend/dist/assets/SoundCloud-qLUb18oY.js +1 -0
  42. chainlit/frontend/dist/assets/Streamable-BvYP7bFp.js +1 -0
  43. chainlit/frontend/dist/assets/Twitch-CTHt-sGZ.js +1 -0
  44. chainlit/frontend/dist/assets/Vidyard-B-0mCJbm.js +1 -0
  45. chainlit/frontend/dist/assets/Vimeo-Dnp7ri8q.js +1 -0
  46. chainlit/frontend/dist/assets/Wistia-DW0x_UBn.js +1 -0
  47. chainlit/frontend/dist/assets/YouTube--98FipvA.js +1 -0
  48. chainlit/frontend/dist/assets/index-D71nZ46o.js +8665 -0
  49. chainlit/frontend/dist/assets/index-g8LTJwwr.css +1 -0
  50. chainlit/frontend/dist/assets/react-plotly-Cn_BQTQw.js +3484 -0
  51. chainlit/frontend/dist/index.html +2 -4
  52. chainlit/haystack/callbacks.py +4 -7
  53. chainlit/input_widget.py +8 -4
  54. chainlit/langchain/callbacks.py +107 -72
  55. chainlit/langflow/__init__.py +1 -0
  56. chainlit/llama_index/__init__.py +2 -2
  57. chainlit/llama_index/callbacks.py +67 -42
  58. chainlit/markdown.py +22 -6
  59. chainlit/message.py +54 -56
  60. chainlit/mistralai/__init__.py +50 -0
  61. chainlit/oauth_providers.py +266 -8
  62. chainlit/openai/__init__.py +10 -18
  63. chainlit/secret.py +1 -1
  64. chainlit/server.py +789 -228
  65. chainlit/session.py +108 -90
  66. chainlit/slack/__init__.py +6 -0
  67. chainlit/slack/app.py +397 -0
  68. chainlit/socket.py +199 -116
  69. chainlit/step.py +141 -89
  70. chainlit/sync.py +2 -1
  71. chainlit/teams/__init__.py +6 -0
  72. chainlit/teams/app.py +338 -0
  73. chainlit/translations/bn.json +235 -0
  74. chainlit/translations/en-US.json +83 -4
  75. chainlit/translations/gu.json +235 -0
  76. chainlit/translations/he-IL.json +235 -0
  77. chainlit/translations/hi.json +235 -0
  78. chainlit/translations/kn.json +235 -0
  79. chainlit/translations/ml.json +235 -0
  80. chainlit/translations/mr.json +235 -0
  81. chainlit/translations/nl-NL.json +233 -0
  82. chainlit/translations/ta.json +235 -0
  83. chainlit/translations/te.json +235 -0
  84. chainlit/translations/zh-CN.json +233 -0
  85. chainlit/translations.py +60 -0
  86. chainlit/types.py +133 -28
  87. chainlit/user.py +14 -3
  88. chainlit/user_session.py +6 -3
  89. chainlit/utils.py +52 -5
  90. chainlit/version.py +3 -2
  91. {chainlit-1.0.400.dist-info → chainlit-2.0.3.dist-info}/METADATA +48 -50
  92. chainlit-2.0.3.dist-info/RECORD +106 -0
  93. chainlit/cli/utils.py +0 -24
  94. chainlit/frontend/dist/assets/index-9711593e.js +0 -723
  95. chainlit/frontend/dist/assets/index-d088547c.css +0 -1
  96. chainlit/frontend/dist/assets/react-plotly-d8762cc2.js +0 -3602
  97. chainlit/playground/__init__.py +0 -2
  98. chainlit/playground/config.py +0 -40
  99. chainlit/playground/provider.py +0 -108
  100. chainlit/playground/providers/__init__.py +0 -13
  101. chainlit/playground/providers/anthropic.py +0 -118
  102. chainlit/playground/providers/huggingface.py +0 -75
  103. chainlit/playground/providers/langchain.py +0 -89
  104. chainlit/playground/providers/openai.py +0 -408
  105. chainlit/playground/providers/vertexai.py +0 -171
  106. chainlit/translations/pt-BR.json +0 -155
  107. chainlit-1.0.400.dist-info/RECORD +0 -66
  108. /chainlit/copilot/dist/assets/{logo_dark-2a3cf740.svg → logo_dark-IkGJ_IwC.svg} +0 -0
  109. /chainlit/copilot/dist/assets/{logo_light-b078e7bc.svg → logo_light-Bb_IPh6r.svg} +0 -0
  110. /chainlit/frontend/dist/assets/{logo_dark-2a3cf740.svg → logo_dark-IkGJ_IwC.svg} +0 -0
  111. /chainlit/frontend/dist/assets/{logo_light-b078e7bc.svg → logo_light-Bb_IPh6r.svg} +0 -0
  112. {chainlit-1.0.400.dist-info → chainlit-2.0.3.dist-info}/WHEEL +0 -0
  113. {chainlit-1.0.400.dist-info → chainlit-2.0.3.dist-info}/entry_points.txt +0 -0
chainlit/frontend/dist/index.html CHANGED
@@ -4,7 +4,6 @@
4
4
  <meta charset="UTF-8" />
5
5
  <meta name="viewport" content="width=device-width, initial-scale=1.0" />
6
6
  <!-- TAG INJECTION PLACEHOLDER -->
7
- <link rel="icon" href="/favicon" />
8
7
  <link rel="preconnect" href="https://fonts.googleapis.com" />
9
8
  <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin />
10
9
  <!-- FONT START -->
@@ -22,11 +21,10 @@
22
21
  <script>
23
22
  const global = globalThis;
24
23
  </script>
25
- <script type="module" crossorigin src="/assets/index-9711593e.js"></script>
26
- <link rel="stylesheet" href="/assets/index-d088547c.css">
24
+ <script type="module" crossorigin src="/assets/index-D71nZ46o.js"></script>
25
+ <link rel="stylesheet" crossorigin href="/assets/index-g8LTJwwr.css">
27
26
  </head>
28
27
  <body>
29
28
  <div id="root"></div>
30
-
31
29
  </body>
32
30
  </html>
chainlit/haystack/callbacks.py CHANGED
@@ -1,14 +1,13 @@
1
1
  import re
2
2
  from typing import Any, Generic, List, Optional, TypeVar
3
3
 
4
- from chainlit.context import context
5
- from chainlit.step import Step
6
- from chainlit.sync import run_sync
7
4
  from haystack.agents import Agent, Tool
8
5
  from haystack.agents.agent_step import AgentStep
9
6
  from literalai.helper import utc_now
10
7
 
11
8
  from chainlit import Message
9
+ from chainlit.step import Step
10
+ from chainlit.sync import run_sync
12
11
 
13
12
  T = TypeVar("T")
14
13
 
@@ -68,9 +67,7 @@ class HaystackAgentCallbackHandler:
68
67
  self.last_tokens: List[str] = []
69
68
  self.answer_reached = False
70
69
 
71
- root_message = context.session.root_message
72
- parent_id = root_message.id if root_message else None
73
- run_step = Step(name=self.agent_name, type="run", parent_id=parent_id)
70
+ run_step = Step(name=self.agent_name, type="run")
74
71
  run_step.start = utc_now()
75
72
  run_step.input = kwargs
76
73
 
@@ -135,7 +132,7 @@ class HaystackAgentCallbackHandler:
135
132
  tool_result: str,
136
133
  tool_name: Optional[str] = None,
137
134
  tool_input: Optional[str] = None,
138
- **kwargs: Any
135
+ **kwargs: Any,
139
136
  ) -> None:
140
137
  # Tool finished, send step with tool_result
141
138
  tool_step = self.stack.pop()
chainlit/input_widget.py CHANGED
@@ -2,8 +2,10 @@ from abc import abstractmethod
2
2
  from collections import defaultdict
3
3
  from typing import Any, Dict, List, Optional
4
4
 
5
+ from pydantic import Field
6
+ from pydantic.dataclasses import dataclass
7
+
5
8
  from chainlit.types import InputWidgetType
6
- from pydantic.dataclasses import Field, dataclass
7
9
 
8
10
 
9
11
  @dataclass
@@ -75,7 +77,7 @@ class Select(InputWidget):
75
77
  initial: Optional[str] = None
76
78
  initial_index: Optional[int] = None
77
79
  initial_value: Optional[str] = None
78
- values: List[str] = Field(default_factory=lambda: [])
80
+ values: List[str] = Field(default_factory=list)
79
81
  items: Dict[str, str] = Field(default_factory=lambda: defaultdict(dict))
80
82
 
81
83
  def __post_init__(
@@ -127,6 +129,7 @@ class TextInput(InputWidget):
127
129
  type: InputWidgetType = "textinput"
128
130
  initial: Optional[str] = None
129
131
  placeholder: Optional[str] = None
132
+ multiline: bool = False
130
133
 
131
134
  def to_dict(self) -> Dict[str, Any]:
132
135
  return {
@@ -137,6 +140,7 @@ class TextInput(InputWidget):
137
140
  "placeholder": self.placeholder,
138
141
  "tooltip": self.tooltip,
139
142
  "description": self.description,
143
+ "multiline": self.multiline,
140
144
  }
141
145
 
142
146
 
@@ -165,8 +169,8 @@ class Tags(InputWidget):
165
169
  """Useful to create an input for an array of strings."""
166
170
 
167
171
  type: InputWidgetType = "tags"
168
- initial: List[str] = Field(default_factory=lambda: [])
169
- values: List[str] = Field(default_factory=lambda: [])
172
+ initial: List[str] = Field(default_factory=list)
173
+ values: List[str] = Field(default_factory=list)
170
174
 
171
175
  def to_dict(self) -> Dict[str, Any]:
172
176
  return {
chainlit/langchain/callbacks.py CHANGED
@@ -1,19 +1,20 @@
1
1
  import json
2
2
  import time
3
- from typing import Any, Dict, List, Optional, TypedDict, Union
3
+ from typing import Any, Dict, List, Optional, Sequence, Tuple, TypedDict, Union
4
4
  from uuid import UUID
5
5
 
6
- from chainlit.context import context_var
7
- from chainlit.message import Message
8
- from chainlit.step import Step
9
- from langchain.callbacks.tracers.base import BaseTracer
6
+ import pydantic
10
7
  from langchain.callbacks.tracers.schemas import Run
11
8
  from langchain.schema import BaseMessage
12
- from langchain.schema.output import ChatGenerationChunk, GenerationChunk
13
9
  from langchain_core.outputs import ChatGenerationChunk, GenerationChunk
10
+ from langchain_core.tracers.base import AsyncBaseTracer
14
11
  from literalai import ChatGeneration, CompletionGeneration, GenerationMessage
15
12
  from literalai.helper import utc_now
16
- from literalai.step import TrueStepType
13
+ from literalai.observability.step import TrueStepType
14
+
15
+ from chainlit.context import context_var
16
+ from chainlit.message import Message
17
+ from chainlit.step import Step
17
18
 
18
19
  DEFAULT_ANSWER_PREFIX_TOKENS = ["Final", "Answer", ":"]
19
20
 
@@ -123,6 +124,14 @@ class GenerationHelper:
123
124
  key: self.ensure_values_serializable(value)
124
125
  for key, value in data.items()
125
126
  }
127
+ elif isinstance(data, pydantic.BaseModel):
128
+ # Fallback to support pydantic v1
129
+ # https://docs.pydantic.dev/latest/migration/#changes-to-pydanticbasemodel
130
+ if pydantic.VERSION.startswith("1"):
131
+ return data.dict()
132
+
133
+ # pydantic v2
134
+ return data.model_dump() # pyright: ignore reportAttributeAccessIssue
126
135
  elif isinstance(data, list):
127
136
  return [self.ensure_values_serializable(item) for item in data]
128
137
  elif isinstance(data, (str, int, float, bool, type(None))):
@@ -229,11 +238,28 @@ class GenerationHelper:
229
238
  return provider, model, tools, settings
230
239
 
231
240
 
232
- DEFAULT_TO_IGNORE = ["Runnable", "<lambda>"]
241
+ def process_content(content: Any) -> Tuple[Dict, Optional[str]]:
242
+ if content is None:
243
+ return {}, None
244
+ if isinstance(content, dict):
245
+ return content, "json"
246
+ elif isinstance(content, str):
247
+ return {"content": content}, "text"
248
+ else:
249
+ return {"content": str(content)}, "text"
250
+
251
+
252
+ DEFAULT_TO_IGNORE = [
253
+ "RunnableSequence",
254
+ "RunnableParallel",
255
+ "RunnableAssign",
256
+ "RunnableLambda",
257
+ "<lambda>",
258
+ ]
233
259
  DEFAULT_TO_KEEP = ["retriever", "llm", "agent", "chain", "tool"]
234
260
 
235
261
 
236
- class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
262
+ class LangchainTracer(AsyncBaseTracer, GenerationHelper, FinalStreamHelper):
237
263
  steps: Dict[str, Step]
238
264
  parent_id_map: Dict[str, str]
239
265
  ignored_runs: set
@@ -252,7 +278,7 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
252
278
  to_keep: Optional[List[str]] = None,
253
279
  **kwargs: Any,
254
280
  ) -> None:
255
- BaseTracer.__init__(self, **kwargs)
281
+ AsyncBaseTracer.__init__(self, **kwargs)
256
282
  GenerationHelper.__init__(self)
257
283
  FinalStreamHelper.__init__(
258
284
  self,
@@ -267,8 +293,6 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
267
293
 
268
294
  if self.context.current_step:
269
295
  self.root_parent_id = self.context.current_step.id
270
- elif self.context.session.root_message:
271
- self.root_parent_id = self.context.session.root_message.id
272
296
  else:
273
297
  self.root_parent_id = None
274
298
 
@@ -282,7 +306,7 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
282
306
  else:
283
307
  self.to_keep = to_keep
284
308
 
285
- def on_chat_model_start(
309
+ async def on_chat_model_start(
286
310
  self,
287
311
  serialized: Dict[str, Any],
288
312
  messages: List[List[BaseMessage]],
@@ -291,8 +315,9 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
291
315
  parent_run_id: Optional["UUID"] = None,
292
316
  tags: Optional[List[str]] = None,
293
317
  metadata: Optional[Dict[str, Any]] = None,
318
+ name: Optional[str] = None,
294
319
  **kwargs: Any,
295
- ) -> Any:
320
+ ) -> Run:
296
321
  lc_messages = messages[0]
297
322
  self.chat_generations[str(run_id)] = {
298
323
  "input_messages": lc_messages,
@@ -301,46 +326,48 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
301
326
  "tt_first_token": None,
302
327
  }
303
328
 
304
- return super().on_chat_model_start(
329
+ return await super().on_chat_model_start(
305
330
  serialized,
306
331
  messages,
307
332
  run_id=run_id,
308
333
  parent_run_id=parent_run_id,
309
334
  tags=tags,
310
335
  metadata=metadata,
336
+ name=name,
311
337
  **kwargs,
312
338
  )
313
339
 
314
- def on_llm_start(
340
+ async def on_llm_start(
315
341
  self,
316
342
  serialized: Dict[str, Any],
317
343
  prompts: List[str],
318
344
  *,
319
345
  run_id: "UUID",
346
+ parent_run_id: Optional[UUID] = None,
320
347
  tags: Optional[List[str]] = None,
321
- parent_run_id: Optional["UUID"] = None,
322
348
  metadata: Optional[Dict[str, Any]] = None,
323
- name: Optional[str] = None,
324
349
  **kwargs: Any,
325
- ) -> Run:
326
- self.completion_generations[str(run_id)] = {
327
- "prompt": prompts[0],
328
- "start": time.time(),
329
- "token_count": 0,
330
- "tt_first_token": None,
331
- }
332
- return super().on_llm_start(
350
+ ) -> None:
351
+ await super().on_llm_start(
333
352
  serialized,
334
353
  prompts,
335
354
  run_id=run_id,
336
355
  parent_run_id=parent_run_id,
337
356
  tags=tags,
338
357
  metadata=metadata,
339
- name=name,
340
358
  **kwargs,
341
359
  )
342
360
 
343
- def on_llm_new_token(
361
+ self.completion_generations[str(run_id)] = {
362
+ "prompt": prompts[0],
363
+ "start": time.time(),
364
+ "token_count": 0,
365
+ "tt_first_token": None,
366
+ }
367
+
368
+ return None
369
+
370
+ async def on_llm_new_token(
344
371
  self,
345
372
  token: str,
346
373
  *,
@@ -348,7 +375,14 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
348
375
  run_id: "UUID",
349
376
  parent_run_id: Optional["UUID"] = None,
350
377
  **kwargs: Any,
351
- ) -> Run:
378
+ ) -> None:
379
+ await super().on_llm_new_token(
380
+ token=token,
381
+ chunk=chunk,
382
+ run_id=run_id,
383
+ parent_run_id=parent_run_id,
384
+ **kwargs,
385
+ )
352
386
  if isinstance(chunk, ChatGenerationChunk):
353
387
  start = self.chat_generations[str(run_id)]
354
388
  else:
@@ -363,24 +397,13 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
363
397
  if self.answer_reached:
364
398
  if not self.final_stream:
365
399
  self.final_stream = Message(content="")
366
- self._run_sync(self.final_stream.send())
367
- self._run_sync(self.final_stream.stream_token(token))
400
+ await self.final_stream.send()
401
+ await self.final_stream.stream_token(token)
368
402
  self.has_streamed_final_answer = True
369
403
  else:
370
404
  self.answer_reached = self._check_if_answer_reached()
371
405
 
372
- return super().on_llm_new_token(
373
- token,
374
- chunk=chunk,
375
- run_id=run_id,
376
- parent_run_id=parent_run_id,
377
- )
378
-
379
- def _run_sync(self, co): # TODO: WHAT TO DO WITH THIS?
380
- context_var.set(self.context)
381
- self.context.loop.create_task(co)
382
-
383
- def _persist_run(self, run: Run) -> None:
406
+ async def _persist_run(self, run: Run) -> None:
384
407
  pass
385
408
 
386
409
  def _get_run_parent_id(self, run: Run):
@@ -431,11 +454,8 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
431
454
  self.ignored_runs.add(str(run.id))
432
455
  return ignore, parent_id
433
456
 
434
- def _is_annotable(self, run: Run):
435
- return run.run_type in ["retriever", "llm"]
436
-
437
- def _start_trace(self, run: Run) -> None:
438
- super()._start_trace(run)
457
+ async def _start_trace(self, run: Run) -> None:
458
+ await super()._start_trace(run)
439
459
  context_var.set(self.context)
440
460
 
441
461
  ignore, parent_id = self._should_ignore_run(run)
@@ -448,40 +468,39 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
448
468
  if ignore:
449
469
  return
450
470
 
451
- step_type: "TrueStepType" = "undefined"
471
+ step_type: TrueStepType = "undefined"
452
472
  if run.run_type == "agent":
453
473
  step_type = "run"
454
474
  elif run.run_type == "chain":
455
- pass
475
+ if not self.steps:
476
+ step_type = "run"
456
477
  elif run.run_type == "llm":
457
478
  step_type = "llm"
458
479
  elif run.run_type == "retriever":
459
- step_type = "retrieval"
480
+ step_type = "tool"
460
481
  elif run.run_type == "tool":
461
482
  step_type = "tool"
462
483
  elif run.run_type == "embedding":
463
484
  step_type = "embedding"
464
485
 
465
- if not self.steps:
466
- step_type = "run"
467
-
468
- disable_feedback = not self._is_annotable(run)
469
-
470
486
  step = Step(
471
487
  id=str(run.id),
472
488
  name=run.name,
473
489
  type=step_type,
474
490
  parent_id=parent_id,
475
- disable_feedback=disable_feedback,
476
491
  )
477
492
  step.start = utc_now()
478
- step.input = run.inputs
493
+ step.input, language = process_content(run.inputs)
494
+ if language is not None:
495
+ if step.metadata is None:
496
+ step.metadata = {}
497
+ step.metadata["language"] = language
479
498
 
480
499
  self.steps[str(run.id)] = step
481
500
 
482
- self._run_sync(step.send())
501
+ await step.send()
483
502
 
484
- def _on_run_update(self, run: Run) -> None:
503
+ async def _on_run_update(self, run: Run) -> None:
485
504
  """Process a run upon update."""
486
505
  context_var.set(self.context)
487
506
 
@@ -499,6 +518,9 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
499
518
  generations = (run.outputs or {}).get("generations", [])
500
519
  generation = generations[0][0]
501
520
  variables = self.generation_inputs.get(str(run.parent_run_id), {})
521
+ variables = {
522
+ k: process_content(v) for k, v in variables.items() if v is not None
523
+ }
502
524
  if message := generation.get("message"):
503
525
  chat_start = self.chat_generations[str(run.id)]
504
526
  duration = time.time() - chat_start["start"]
@@ -525,13 +547,21 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
525
547
  # find first message with prompt_id
526
548
  for m in chat_start["input_messages"]:
527
549
  if m.additional_kwargs.get("prompt_id"):
528
- current_step.generation.prompt_id = m.additional_kwargs["prompt_id"]
550
+ current_step.generation.prompt_id = m.additional_kwargs[
551
+ "prompt_id"
552
+ ]
529
553
  if custom_variables := m.additional_kwargs.get("variables"):
530
- current_step.generation.variables = custom_variables
554
+ current_step.generation.variables = {
555
+ k: process_content(v)
556
+ for k, v in custom_variables.items()
557
+ if v is not None
558
+ }
531
559
  break
532
-
560
+
533
561
  current_step.language = "json"
534
- current_step.output = json.dumps(message_completion)
562
+ current_step.output = json.dumps(
563
+ message_completion, indent=4, ensure_ascii=False
564
+ )
535
565
  else:
536
566
  completion_start = self.completion_generations[str(run.id)]
537
567
  completion = generation.get("text", "")
@@ -555,34 +585,39 @@ class LangchainTracer(BaseTracer, GenerationHelper, FinalStreamHelper):
555
585
 
556
586
  if current_step:
557
587
  current_step.end = utc_now()
558
- self._run_sync(current_step.update())
588
+ await current_step.update()
559
589
 
560
590
  if self.final_stream and self.has_streamed_final_answer:
561
- if self.final_stream.content:
562
- self.final_stream.content = completion
563
- self._run_sync(self.final_stream.update())
591
+ await self.final_stream.update()
564
592
 
565
593
  return
566
594
 
567
595
  outputs = run.outputs or {}
568
596
  output_keys = list(outputs.keys())
569
597
  output = outputs
598
+
570
599
  if output_keys:
571
600
  output = outputs.get(output_keys[0], outputs)
572
601
 
573
602
  if current_step:
574
- current_step.output = output
603
+ current_step.output = (
604
+ output[0]
605
+ if isinstance(output, Sequence)
606
+ and not isinstance(output, str)
607
+ and len(output)
608
+ else output
609
+ )
575
610
  current_step.end = utc_now()
576
- self._run_sync(current_step.update())
611
+ await current_step.update()
577
612
 
578
- def _on_error(self, error: BaseException, *, run_id: UUID, **kwargs: Any):
613
+ async def _on_error(self, error: BaseException, *, run_id: UUID, **kwargs: Any):
579
614
  context_var.set(self.context)
580
615
 
581
616
  if current_step := self.steps.get(str(run_id), None):
582
617
  current_step.is_error = True
583
618
  current_step.output = str(error)
584
619
  current_step.end = utc_now()
585
- self._run_sync(current_step.update())
620
+ await current_step.update()
586
621
 
587
622
  on_llm_error = _on_error
588
623
  on_chain_error = _on_error
chainlit/langflow/__init__.py CHANGED
@@ -8,6 +8,7 @@ if not check_module_version("langflow", "0.1.4"):
8
8
  from typing import Dict, Optional, Union
9
9
 
10
10
  import httpx
11
+
11
12
  from chainlit.telemetry import trace_event
12
13
 
13
14
 
chainlit/llama_index/__init__.py CHANGED
@@ -1,6 +1,6 @@
1
1
  from chainlit.utils import check_module_version
2
2
 
3
- if not check_module_version("llama_index", "0.8.3"):
3
+ if not check_module_version("llama_index.core", "0.10.15"):
4
4
  raise ValueError(
5
- "Expected LlamaIndex version >= 0.8.3. Run `pip install llama_index --upgrade`"
5
+ "Expected LlamaIndex version >= 0.10.15. Run `pip install llama_index --upgrade`"
6
6
  )