lalamo 0.5.6__tar.gz → 0.5.7__tar.gz

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 (103) hide show
  1. {lalamo-0.5.6 → lalamo-0.5.7}/PKG-INFO +1 -1
  2. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/__init__.py +1 -1
  3. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/speculator/estimator.py +2 -1
  4. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/speculator/inference.py +3 -1
  5. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo.egg-info/PKG-INFO +1 -1
  6. {lalamo-0.5.6 → lalamo-0.5.7}/LICENSE +0 -0
  7. {lalamo-0.5.6 → lalamo-0.5.7}/README.md +0 -0
  8. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/common.py +0 -0
  9. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/data/__init__.py +0 -0
  10. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/data/huggingface_message.py +0 -0
  11. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/data/lalamo_completions.py +0 -0
  12. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/data/utils.py +0 -0
  13. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/main.py +0 -0
  14. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/message_processor.py +0 -0
  15. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/__init__.py +0 -0
  16. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/common.py +0 -0
  17. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/__init__.py +0 -0
  18. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/common.py +0 -0
  19. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/executorch.py +0 -0
  20. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/__init__.py +0 -0
  21. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/common.py +0 -0
  22. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/gemma2.py +0 -0
  23. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/gemma3.py +0 -0
  24. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/gpt_oss.py +0 -0
  25. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/llama.py +0 -0
  26. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/llamba.py +0 -0
  27. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/mistral.py +0 -0
  28. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/modern_bert.py +0 -0
  29. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/qwen2.py +0 -0
  30. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/decoder_configs/huggingface/qwen3.py +0 -0
  31. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/huggingface_generation_config.py +0 -0
  32. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/huggingface_tokenizer_config.py +0 -0
  33. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/loaders/__init__.py +0 -0
  34. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/loaders/common.py +0 -0
  35. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/loaders/executorch.py +0 -0
  36. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/loaders/huggingface.py +0 -0
  37. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/loaders/utils.py +0 -0
  38. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/__init__.py +0 -0
  39. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/common.py +0 -0
  40. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/deepseek.py +0 -0
  41. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/gemma.py +0 -0
  42. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/gpt_oss.py +0 -0
  43. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/huggingface.py +0 -0
  44. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/llama.py +0 -0
  45. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/llamba.py +0 -0
  46. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/mirai.py +0 -0
  47. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/mistral.py +0 -0
  48. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/pleias.py +0 -0
  49. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/polaris.py +0 -0
  50. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/qwen.py +0 -0
  51. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/model_import/model_specs/reka.py +0 -0
  52. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/models/__init__.py +0 -0
  53. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/models/common.py +0 -0
  54. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/models/language_model.py +0 -0
  55. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/models/router.py +0 -0
  56. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/__init__.py +0 -0
  57. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/activations.py +0 -0
  58. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/classifier.py +0 -0
  59. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/common.py +0 -0
  60. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/decoder.py +0 -0
  61. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/embedding.py +0 -0
  62. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/linear.py +0 -0
  63. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/mlp.py +0 -0
  64. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/mlx_interop.py +0 -0
  65. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/normalization.py +0 -0
  66. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/rope.py +0 -0
  67. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/token_mixers/__init__.py +0 -0
  68. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/token_mixers/attention.py +0 -0
  69. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/token_mixers/common.py +0 -0
  70. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/token_mixers/mamba.py +0 -0
  71. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/token_mixers/state/__init__.py +0 -0
  72. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/token_mixers/state/common.py +0 -0
  73. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/token_mixers/state/kv_cache.py +0 -0
  74. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/token_mixers/state/mamba_state.py +0 -0
  75. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/torch_interop.py +0 -0
  76. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/transformer.py +0 -0
  77. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/transformer_layer.py +0 -0
  78. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/modules/utils.py +0 -0
  79. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/quantization.py +0 -0
  80. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/registry_abc.py +0 -0
  81. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/sampling.py +0 -0
  82. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/speculator/__init__.py +0 -0
  83. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/speculator/common.py +0 -0
  84. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/speculator/ngram.py +0 -0
  85. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/speculator/utils.py +0 -0
  86. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo/utils.py +0 -0
  87. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo.egg-info/SOURCES.txt +0 -0
  88. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo.egg-info/dependency_links.txt +0 -0
  89. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo.egg-info/entry_points.txt +0 -0
  90. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo.egg-info/requires.txt +0 -0
  91. {lalamo-0.5.6 → lalamo-0.5.7}/lalamo.egg-info/top_level.txt +0 -0
  92. {lalamo-0.5.6 → lalamo-0.5.7}/pyproject.toml +0 -0
  93. {lalamo-0.5.6 → lalamo-0.5.7}/setup.cfg +0 -0
  94. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_cartesia_mlx_models.py +0 -0
  95. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_generation.py +0 -0
  96. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_huggingface_model_conversion.py +0 -0
  97. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_huggingface_models.py +0 -0
  98. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_mlx_models.py +0 -0
  99. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_model_spec.py +0 -0
  100. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_models.py +0 -0
  101. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_moe.py +0 -0
  102. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_parameter_tree.py +0 -0
  103. {lalamo-0.5.6 → lalamo-0.5.7}/tests/test_registry_abc.py +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: lalamo
3
- Version: 0.5.6
3
+ Version: 0.5.7
4
4
  Summary: JAX library for optimization and export of models for use with the UZU inference engine.
5
5
  Requires-Python: <4,>=3.12
6
6
  Description-Content-Type: text/markdown
@@ -15,7 +15,7 @@ from lalamo.speculator import (
15
15
  SpeculatorTrainingEvent,
16
16
  )
17
17
 
18
- __version__ = "0.5.6"
18
+ __version__ = "0.5.7"
19
19
 
20
20
  __all__ = [
21
21
  "AssistantMessage",
@@ -19,13 +19,14 @@ def estimate_memory_from_batchsize(
19
19
  memory_analysis = (
20
20
  jax.jit(
21
21
  functools.partial(
22
- model.generate_tokens,
22
+ LanguageModel.generate_tokens,
23
23
  max_output_length=max_output_length,
24
24
  num_top_logits_to_return=num_logits_per_token,
25
25
  ),
26
26
  backend="cpu", # cuda backend tries to allocate in .compile() and ooms
27
27
  )
28
28
  .lower(
29
+ model,
29
30
  prompt_token_ids=jax.ShapeDtypeStruct((batch_size, max_input_length), jnp.int32),
30
31
  prompt_lengths_without_padding=jax.ShapeDtypeStruct((batch_size,), jnp.int32),
31
32
  )
@@ -30,12 +30,13 @@ def inference_collect_traces(
30
30
  generate_tokens_compiled = (
31
31
  jax.jit(
32
32
  functools.partial(
33
- model.generate_tokens,
33
+ LanguageModel.generate_tokens,
34
34
  max_output_length=max_output_length,
35
35
  num_top_logits_to_return=num_top_logits_to_collect,
36
36
  ),
37
37
  )
38
38
  .lower(
39
+ model,
39
40
  prompt_token_ids=jax.ShapeDtypeStruct((batch_size, max_input_length), jnp.int32),
40
41
  prompt_lengths_without_padding=jax.ShapeDtypeStruct((batch_size,), jnp.int32),
41
42
  )
@@ -60,6 +61,7 @@ def inference_collect_traces(
60
61
  )
61
62
 
62
63
  generated = generate_tokens_compiled(
64
+ model,
63
65
  prompt_token_ids=padded,
64
66
  prompt_lengths_without_padding=length_without_padding,
65
67
  )
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: lalamo
3
- Version: 0.5.6
3
+ Version: 0.5.7
4
4
  Summary: JAX library for optimization and export of models for use with the UZU inference engine.
5
5
  Requires-Python: <4,>=3.12
6
6
  Description-Content-Type: text/markdown
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes
File without changes