instruction stringclasses 100
values | code stringlengths 78 193k | response stringlengths 259 170k | file stringlengths 59 203 |
|---|---|---|---|
Provide docstrings following PEP 257 |
import functools
from typing import Literal
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .bert_model import BertConfig, BertModel
def huggingface(
model_config: BertConfig,
quantization: Quantization,
hf_prefix: Literal["", "bert."] = "... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's BERT parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
from typing import Literal
@@ -15,6 +19,26 @@ quantization: Quantization,
hf_prefix: Literal["", "bert."] = "",
) -> ExternMap... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/bert/bert_loader.py |
Document this module using docstrings |
import dataclasses
import logging
from typing import Any, Dict, Optional
from tvm import tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Module, Tensor
from tvm.relax.frontend.nn.op import permute_dims, reshape, wrap_nested
from tvm.relax.op import strided_slice
from mlc_llm import op as op_e... | --- +++ @@ -1,3 +1,7 @@+"""
+Implementation of LLaVa Model
+Implements the CLIP Vision Encoder. Uses Llama for the Language Encoder.
+"""
import dataclasses
import logging
@@ -30,6 +34,9 @@
@dataclasses.dataclass
class LlavaConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """
+ LLaVa C... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/llava/llava_model.py |
Create documentation strings for testing functions |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .gpt_neox_model import GPTNeoXConfig, GPTNeoXForCausalLM
def huggingface(model_config: GPTNeoXConfig, quantization: Quantization) -> ExternMapping:
model = GPTNeoXForCausalLM(model_c... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's GPTNeoX parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -10,6 +14,22 @@
def huggingface(model_config: GPTNeoXConfig, quantization: Quantization) -> ExternMapping:
+ """Returns a ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/gpt_neox/gpt_neox_loader.py |
Document classes and their methods |
import dataclasses
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm.model.gemma.gemma_model import (
GemmaAttention,
GemmaConfig,
GemmaForCausalLM,
GemmaMLP,
GemmaModel,
)
from mlc_llm.nn import PagedKVCache
from mlc_llm.support import logging
from mlc_l... | --- +++ @@ -1,3 +1,4 @@+"""Implementation for Gemma2 architecture."""
import dataclasses
@@ -20,6 +21,7 @@
@dataclasses.dataclass
class Gemma2Config(GemmaConfig):
+ """Configuration of the Gemma2 model, in addition to the Gemma model"""
# NOTE: We ignore attn_logit_softcapping in the gemma2 implementa... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/gemma2/gemma2_model.py |
Add docstrings that explain logic |
import functools
from typing import Callable, List
import numpy as np
from mlc_llm.loader import ExternMapping, QuantizeMapping
from mlc_llm.quantization import BlockScaleQuantize, Quantization
from .deepseek_v2_model import DeepseekV2Config, DeepseekV2ForCausalLM
def huggingface( # pylint: disable=too-many-loca... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Deepseek-V2 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
from typing import Callable, List
@@ -13,6 +17,22 @@ def huggingface( # pylint: disable=too-many-locals,too-many-statements
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/deepseek_v2/deepseek_v2_loader.py |
Help me write clear docstrings |
import functools
from typing import Callable, List, Optional, Tuple
import numpy as np
from mlc_llm.loader import ExternMapping, QuantizeMapping
from mlc_llm.quantization import BlockScaleQuantize, Quantization
from .ministral3_model import Ministral3Config, Mistral3ForConditionalGeneration
def _dequantize_block_... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Ministral3 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
from typing import Callable, List, Optional, Tuple
@@ -13,6 +17,7 @@ def _dequantize_block_scale_weight( # pylint: disable=too-ma... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/ministral3/ministral3_loader.py |
Generate docstrings for script automation |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.loader.standard_loader import make_standard_hf_loader
from mlc_llm.quantization import Quantization, make_awq_quant
from .llava_model import LlavaConfig, LlavaForCausalLM
awq_quant = make_awq_quant(LlavaForCausalLM)
def _nu... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Llava parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -25,6 +29,21 @@
def awq(model_config: LlavaConfig, quantization: Quantization) -> ExternMapping:
+ """Returns a parameter ma... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/llava/llava_loader.py |
Create simple docstrings for beginners |
import dataclasses
from typing import Any, Dict, Optional
from tvm.relax.frontend import nn
from mlc_llm.support import logging
from mlc_llm.support.config import ConfigBase
logger = logging.getLogger(__name__)
@dataclasses.dataclass
class MedusaConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
... | --- +++ @@ -1,3 +1,4 @@+"""Medusa model definition."""
import dataclasses
from typing import Any, Dict, Optional
@@ -12,6 +13,7 @@
@dataclasses.dataclass
class MedusaConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Llama model."""
medusa_num_heads: int
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/medusa/medusa_model.py |
Create simple docstrings for beginners |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .minicpm_model import MiniCPMConfig, MiniCPMForCausalLM
def huggingface(model_config: MiniCPMConfig, quantization: Quantization) -> ExternMapping:
model = MiniCPMForCausalLM(model_co... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's MiniCPM parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -10,6 +14,22 @@
def huggingface(model_config: MiniCPMConfig, quantization: Quantization) -> ExternMapping:
+ """Returns a ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/minicpm/minicpm_loader.py |
Create docstrings for all classes and functions |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor_parall... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Orion-14B architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class OrionConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Orion model."""
hidden... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/orion/orion_model.py |
Insert docstrings into my code |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.model.gemma.gemma_model import GemmaEmbedding
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.suppo... | --- +++ @@ -1,3 +1,4 @@+"""Implementation for Gemma3 architecture."""
import dataclasses
from typing import Any, Dict, Optional
@@ -19,6 +20,7 @@
@dataclasses.dataclass
class Gemma3TextConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the text model inside Gemma3"""
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/gemma3/gemma3_model.py |
Create documentation strings for testing functions |
import dataclasses
import math
from functools import partial
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.nn.expert import Mixtr... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Minicpm architecture.
+"""
import dataclasses
import math
@@ -21,6 +24,7 @@
@dataclasses.dataclass
class MiniCPMConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the MiniCPM model."""
vocab_size: int
hidden_siz... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/minicpm/minicpm_model.py |
Generate helpful docstrings for debugging |
import dataclasses
from typing import Any, Dict, Optional, Union
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Phi architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional, Union
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class Phi1Config(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Phi-1/Phi-1.5 model."""
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/phi/phi_model.py |
Create documentation for each function signature |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor_parall... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for CHATGLM3 architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class GLMConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the ChatGLM model."""
hidden_... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/chatglm3/chatglm3_model.py |
Add structured docstrings to improve clarity |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .phi_model import Phi1Config, PhiConfig, PhiForCausalLM
def huggingface(model_config: PhiConfig, quantization: Quantization) -> ExternMapping:
model = PhiForCausalLM(model_config)
... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Phi parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -10,6 +14,22 @@
def huggingface(model_config: PhiConfig, quantization: Quantization) -> ExternMapping:
+ """Returns a paramete... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/phi/phi_loader.py |
Add docstrings with type hints explained | # pylint: disable=W0611
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .internlm2_model import InternLM2Config, InternLM2ForCausalLM
def huggingface(model_config: InternLM2ForCausalLM, quantization: Quantization) -> ExternMapping:
... | --- +++ @@ -1,4 +1,8 @@ # pylint: disable=W0611
+"""
+This file specifies how MLC's InternLM2 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -11,6 +15,22 @@
def huggingface(model_config: InternLM2ForCausalLM, quantization: Quantization) ->... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/internlm2/internlm2_loader.py |
Create simple docstrings for beginners |
import dataclasses
from functools import partial
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_ll... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for QWEN2 architecture.
+"""
import dataclasses
from functools import partial
@@ -19,6 +22,7 @@
@dataclasses.dataclass
class QWen2Config(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the QWen2 model."""
hidden_act: str
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen2/qwen2_model.py |
Generate docstrings with parameter types |
import functools
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .gpt2_model import GPT2Config, GPT2LMHeadModel
def huggingface(model_config: GPT2Config, quantization: Quantization) -> ExternMapping:
model = GPT2LMHeadModel(model_config)
if quantization is not No... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's GPT-2 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -8,6 +12,22 @@
def huggingface(model_config: GPT2Config, quantization: Quantization) -> ExternMapping:
+ """Returns a parame... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/gpt2/gpt2_loader.py |
Generate docstrings for this script |
import dataclasses
import math
from functools import partial
from typing import Any, Dict, Optional, Tuple
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import ... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Ministral 3 architecture.
+"""
import dataclasses
import math
@@ -20,6 +23,7 @@
@dataclasses.dataclass
class Ministral3Config(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Ministral 3 model."""
hidden_size: int
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/ministral3/ministral3_model.py |
Document this script properly |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor_parall... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Mistral architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class MistralConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Mistral model."""
hidd... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/mistral/mistral_model.py |
Add detailed documentation for each class |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .mixtral_model import MixtralConfig, MixtralForCausalLM
def huggingface(model_config: MixtralConfig, quantization: Quantization) -> ExternMapping:
model = MixtralForCausalLM(model_co... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Mixtral parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -10,6 +14,22 @@
def huggingface(model_config: MixtralConfig, quantization: Quantization) -> ExternMapping:
+ """Returns a ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/mixtral/mixtral_loader.py |
Document helper functions with docstrings |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.loader.standard_loader import make_standard_hf_loader
from mlc_llm.quantization import Quantization, make_awq_quant
from .mistral_model import MistralConfig, MistralForCausalLM
awq_quant = make_awq_quant(MistralForCausalLM)
... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Mistral parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -19,6 +23,21 @@
def awq(model_config: MistralConfig, quantization: Quantization) -> ExternMapping:
+ """Returns a paramete... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/mistral/mistral_loader.py |
Generate NumPy-style docstrings |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .qwen2_moe_model import Qwen2MoeConfig, Qwen2MoeForCausalLM
def huggingface(model_config: Qwen2MoeConfig, quantization: Quantization) -> ExternMapping:
model = Qwen2MoeForCausalLM(mo... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's QWen2 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -10,6 +14,22 @@
def huggingface(model_config: Qwen2MoeConfig, quantization: Quantization) -> ExternMapping:
+ """Returns a p... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen2_moe/qwen2_moe_loader.py |
Replace inline comments with docstrings |
# pylint: disable=missing-function-docstring,missing-class-docstring
import dataclasses
from functools import partial
from typing import Any, Dict, Optional, Tuple
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn ... | --- +++ @@ -1,3 +1,18 @@+"""Partial Qwen2.5-VL implementation focused on decoder-side MRoPE support.
+
+This file intentionally does not provide complete end-to-end Qwen2.5-VL support yet.
+The current scope is:
+
+- decoder-side text model structure,
+- multimodal rotary embedding generation/application,
+- position-i... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen2_5_vl/qwen2_5_vl_model.py |
Generate consistent documentation across files |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor_parall... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for QWEN architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class QWenConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the QWen model."""
vocab_size: i... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen/qwen_model.py |
Generate docstrings with parameter types |
import dataclasses
from tvm import tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.model.llama.llama_model import (
LlamaAttention,
LlamaConfig,
LlamaForCausalLM,
LlamaModel,
)
from mlc_llm.nn import PagedKVCache
from ml... | --- +++ @@ -1,3 +1,4 @@+"""Implementation for Mistral architecture."""
import dataclasses
@@ -22,6 +23,7 @@
@dataclasses.dataclass
class MixtralConfig(LlamaConfig): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Mixtral model."""
num_local_experts: int = 0
num_experts_p... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/mixtral/mixtral_model.py |
Add structured docstrings to improve clarity |
import dataclasses
from typing import Any, Dict, Optional
from tvm import relax, target, te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.model.phi3 import Phi3Model
from mlc_llm.model.vision import CLIPVisionConfig, ImageProcessor
f... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Phi architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -33,6 +36,7 @@
@dataclasses.dataclass
class Phi3VConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Phi-3 Vision model."""
model... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/phi3v/phi3v_model.py |
Include argument descriptions in docstrings |
import dataclasses
from typing import Any, Callable, Dict, Literal, Optional, Tuple
from tvm.relax.frontend import nn
from mlc_llm.loader import ExternMapping, QuantizeMapping
from mlc_llm.quantization import make_quantization_functions
from mlc_llm.quantization.quantization import Quantization
from .baichuan impor... | --- +++ @@ -1,3 +1,4 @@+"""A centralized registry of all existing model architures and their configurations."""
import dataclasses
from typing import Any, Callable, Dict, Literal, Optional, Tuple
@@ -62,6 +63,19 @@
@dataclasses.dataclass
class EmbeddingMetadata:
+ """Embedding model metadata.
+
+ Parameter... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/model.py |
Insert docstrings into my code |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor_parall... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for StableLM architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class StableLmConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the StableLM model."""
v... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/stable_lm/stablelm_model.py |
Add structured docstrings to improve clarity |
import dataclasses
from typing import Any, Dict, Optional, Tuple
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Object, Tensor, op
from tvm.script import tir as T
from mlc_llm.nn.rnn_state import RNNState
from mlc_llm.support import logging
from mlc_llm.support.config imp... | --- +++ @@ -1,3 +1,4 @@+"""Implementation for RWKV6 architecture."""
import dataclasses
from typing import Any, Dict, Optional, Tuple
@@ -16,6 +17,7 @@
@dataclasses.dataclass
class StateID:
+ """State ID for RWKV6."""
ATT_X = 0
ATT_KV = 1
@@ -24,6 +26,7 @@
@dataclasses.dataclass
class RWKV6Conf... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/rwkv6/rwkv6_model.py |
Turn comments into proper docstrings |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Object, Tensor, op
from tvm.script import tir as T
from mlc_llm.nn.rnn_state import RNNState
from mlc_llm.support import logging
from mlc_llm.support.config import Con... | --- +++ @@ -1,3 +1,4 @@+"""Implementation for RWKV5 architecture."""
import dataclasses
from typing import Any, Dict, Optional
@@ -16,6 +17,7 @@
@dataclasses.dataclass
class StateID:
+ """State ID for RWKV5."""
ATT_X = 0
ATT_KV = 1
@@ -24,6 +26,7 @@
@dataclasses.dataclass
class RWKV5Config(Conf... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/rwkv5/rwkv5_model.py |
Add structured docstrings to improve clarity |
import functools
from ...loader import ExternMapping
from ...quantization import Quantization
from .rwkv6_model import RWKV6_ForCausalLM, RWKV6Config
def huggingface(model_config: RWKV6Config, quantization: Quantization) -> ExternMapping:
model = RWKV6_ForCausalLM(model_config)
if quantization is not None:
... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's RWKV6 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -7,6 +11,22 @@
def huggingface(model_config: RWKV6Config, quantization: Quantization) -> ExternMapping:
+ """Returns a param... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/rwkv6/rwkv6_loader.py |
Create documentation for each function signature |
from tvm import relax, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Module, Tensor, op
from tvm.script import tir as T
from mlc_llm.model.vision import CLIPVisionModel
from mlc_llm.support.config import ConfigBase
# mypy: disable-error-code="attr-defined"
# pylint: disable=invalid-name,mi... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Phi architecture.
+"""
from tvm import relax, tir
from tvm.relax.frontend import nn
@@ -262,6 +265,10 @@ return image_features_hd
def add_image_newline(self, image_features_hd):
+ """
+ image_features_hd: (num_images, h_crop*12, w_crop*12,... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/phi3v/phi3v_image.py |
Write clean docstrings for readability |
import dataclasses
import logging
from typing import Any, Dict, Tuple
from tvm import relax
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Module, Tensor
from tvm.relax.frontend.nn.modules import Conv2D
from tvm.relax.frontend.nn.op import (
add,
broadcast_to,
concat,
permute_dims... | --- +++ @@ -1,3 +1,6 @@+"""
+Implements the CLIP Vision Encoder.
+"""
import dataclasses
import logging
@@ -25,6 +28,9 @@
@dataclasses.dataclass
class CLIPVisionConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """
+ Config for the vision encoder
+ """
hidden_size: int
im... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/vision/clip_vision.py |
Document classes and their methods |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor_parall... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Starcoder2 architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class Starcoder2Config(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Starcoder2 model."""
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/starcoder2/starcoder2_model.py |
Help me document legacy Python code |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .starcoder2_model import Starcoder2Config, Starcoder2ForCausalLM
def huggingface(model_config: Starcoder2Config, quantization: Quantization) -> ExternMapping:
model = Starcoder2ForCa... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Starcoder2 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -10,6 +14,22 @@
def huggingface(model_config: Starcoder2Config, quantization: Quantization) -> ExternMapping:
+ """Retu... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/starcoder2/starcoder2_loader.py |
Add docstrings to improve readability |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor_parall... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Phi-3 architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class Phi3Config(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Phi-3 model."""
model_type:... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/phi3/phi3_model.py |
Add well-formatted docstrings |
from typing import Sequence, Union
from tvm import relax as rx
from tvm import tir
from tvm.relax.frontend.nn import Object, Tensor
from tvm.script import tir as T
class RNNState(Object):
@staticmethod
def create(
max_batch_size: tir.Var,
num_hidden_layers: int,
max_history: int,
... | --- +++ @@ -1,3 +1,4 @@+"""RNN State modeling."""
from typing import Sequence, Union
@@ -8,6 +9,7 @@
class RNNState(Object):
+ """The RNN State used in Space State Models"""
@staticmethod
def create(
@@ -17,6 +19,19 @@ init_values: Sequence[Tensor],
name: str = "rnn_state",
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/nn/rnn_state.py |
Generate docstrings for this script |
import tvm
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm.support import logging
from . import extern as _extern
logger = logging.getLogger(__name__)
WARN_FLASHINFER_GROUP_SIZE = False
WARN_FLASHINFER_HEAD_DIM = False
def attention( # pylint: disable=invalid-name,to... | --- +++ @@ -1,3 +1,4 @@+"""Operators enabled by external modules."""
import tvm
from tvm.relax.frontend import nn
@@ -22,6 +23,40 @@ attn_score_scaling_factor: float = 1.0,
qk_dtype: str = None,
) -> nn.Tensor:
+ """Attention with casual mask.
+
+ --- Variables ---
+ s: sequence length of the cur... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/attention.py |
Add concise docstrings to each method |
import dataclasses
from functools import partial
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_ll... | --- +++ @@ -1,3 +1,7 @@+"""
+Implementation for OLMo architecture.
+TODO: add docstring
+"""
import dataclasses
from functools import partial
@@ -19,6 +23,7 @@
@dataclasses.dataclass
class OLMoConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the OLMo model."""
v... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/olmo/olmo_model.py |
Add docstrings explaining edge cases |
import dataclasses
from typing import Optional
from tvm.target import Target
@dataclasses.dataclass
class ExternModuleStore:
configured: bool = False
target: Optional[Target] = None
flashinfer: bool = False
faster_transformer: bool = False
cutlass_group_gemm: bool = False
cutlass_gemm: bool... | --- +++ @@ -1,3 +1,19 @@+"""Potential externel modules managed by MLC compilation stack.
+
+An externl module could contain one or multiple handcrafted kernels, as long as it is provided as
+an object file (`.o`), a C++ source file (`.cc`), or a CUDA source file (`.cu`). It can be
+integrated into the system pretty smo... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/extern.py |
Add professional docstrings to my codebase |
import functools
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .phi3_model import Phi3Config, Phi3ForCausalLM
def phi3_huggingface(model_config: Phi3Config, quantization: Quantization) -> ExternMapping:
model = Phi3ForCausalLM(model_config)
if quantization is n... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Phi parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -8,6 +12,22 @@
def phi3_huggingface(model_config: Phi3Config, quantization: Quantization) -> ExternMapping:
+ """Returns a par... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/phi3/phi3_loader.py |
Write docstrings for this repository |
from typing import Literal, Optional, Tuple
from tvm import DataType, DataTypeCode, s_tir, tir
from tvm.relax.frontend.nn import Tensor, op
from tvm.script import tir as T
# mypy: disable-error-code="attr-defined,valid-type,name-defined"
# pylint: disable=too-many-locals,invalid-name,too-many-arguments,too-many-stat... | --- +++ @@ -1,3 +1,4 @@+"""Mixture of Experts operators"""
from typing import Literal, Optional, Tuple
@@ -10,6 +11,29 @@
def gemv(x: Tensor, w: Tensor, indptr: Tensor) -> Tensor:
+ """GEMV for project-in (e1-e3) or project-out (e2) in MLP.
+
+ Parameters
+ ----------
+ x : Tensor
+ For proj... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/moe_matmul.py |
Add docstrings to clarify complex logic |
from tvm.script import tir as T
# mypy: disable-error-code="attr-defined,valid-type,name-defined"
# pylint: disable=too-many-locals,invalid-name,too-many-arguments,
# pylint: disable=too-many-statements,line-too-long,too-many-nested-blocks,too-many-branches
def batch_spec_verify(vocab_size):
TX = 1024
def ... | --- +++ @@ -1,3 +1,4 @@+"""Operators for batch verify in speculative decoding."""
from tvm.script import tir as T
@@ -7,6 +8,49 @@
def batch_spec_verify(vocab_size):
+ """Batch draft verify function. This function verifies the token tree.
+
+ Before calling the function
+
+ - token_tree_parent_ptr[b] ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/batch_spec_verify.py |
Include argument descriptions in docstrings |
import operator
from functools import reduce
from typing import Optional
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import op
def faster_transformer_dequantize_gemm( # pylint: disable=too-many-arguments
x: nn.Tensor,
weight: nn.Tensor,
scale: nn.Tensor,
bias: Optional[nn.Tensor] =... | --- +++ @@ -1,3 +1,4 @@+"""Operators enabled by external modules."""
import operator
from functools import reduce
@@ -15,6 +16,31 @@ activation: Optional[str] = None,
group_size: Optional[int] = None,
):
+ """
+ Faster Transformer dequantize gemm inference with CutlassFpAIntB
+
+ Parameters
+ ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/ft_gemm.py |
Add concise docstrings to each method |
from __future__ import annotations
from dataclasses import dataclass
from typing import List, Optional, Sequence, Tuple
import numpy as np
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
def _rotate_half(x: Tensor) -> Tensor:
x1, x2 = op.split(x, 2, axis=... | --- +++ @@ -1,3 +1,4 @@+"""Utilities for Multimodal Rotary Position Embeddings (MRoPE)."""
from __future__ import annotations
@@ -11,6 +12,7 @@
def _rotate_half(x: Tensor) -> Tensor:
+ """Rotate the last dimension of ``x`` by swapping pairs."""
x1, x2 = op.split(x, 2, axis=-1)
return op.concat([... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/mrope.py |
Generate docstrings with examples |
from typing import Literal, Optional
from pydantic import BaseModel
class DisaggConfig(BaseModel):
kind: Optional[Literal["prepare_receive", "remote_send", "start_generation"]] = None
# "kv_append_metadata" is base64-encoded and is thus a string.
kv_append_metadata: Optional[str] = None
# "kv_windo... | --- +++ @@ -1,3 +1,4 @@+"""Debug protocols in MLC LLM"""
from typing import Literal, Optional
@@ -5,6 +6,7 @@
class DisaggConfig(BaseModel):
+ """The class of metadata used in microserving APIs."""
kind: Optional[Literal["prepare_receive", "remote_send", "start_generation"]] = None
# "kv_append_... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/protocol/debug_protocol.py |
Add docstrings to incomplete code |
import functools
import numpy as np
from mlc_llm.loader import ExternMapping
from mlc_llm.loader.standard_loader import make_standard_hf_loader
from mlc_llm.quantization import Quantization, make_awq_quant
from .olmo_model import OLMoConfig, OLMoForCausalLM
awq_quant = make_awq_quant(OLMoForCausalLM)
huggingface... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's OLMo parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -19,6 +23,21 @@
def awq(model_config: OLMoConfig, quantization: Quantization) -> ExternMapping:
+ """Returns a parameter mapp... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/olmo/olmo_loader.py |
Document all public functions with docstrings |
from pydantic import BaseModel
from mlc_llm.protocol.openai_api_protocol import CompletionRequest
class PrepRecvRequest(CompletionRequest):
end: int
class PrepRecvResponse(BaseModel):
kv_append_metadata: str
prefix_matched_length: int
class RemoteSendRequest(CompletionRequest):
begin: int
... | --- +++ @@ -1,3 +1,4 @@+"""Protocols in MLC LLM for MicroServing."""
from pydantic import BaseModel
@@ -5,17 +6,53 @@
class PrepRecvRequest(CompletionRequest):
+ """The extra request body for prep_recv request in MicroServing.
+
+ Attributes
+ ----------
+ kv_window_end : int
+ [0, kv_window... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/protocol/microserving_protocol.py |
Create docstrings for all classes and functions |
import functools
from typing import Callable, List, Literal
import numpy as np
from mlc_llm.loader import ExternMapping, QuantizeMapping
from mlc_llm.quantization import BlockScaleQuantize, Quantization
from .qwen3_model import Qwen3Config, Qwen3LMHeadModel
def huggingface(
model_config: Qwen3Config,
quan... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's QWen2 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
from typing import Callable, List, Literal
@@ -15,6 +19,26 @@ quantization: Quantization,
hf_prefix: Literal["", "model."] = "... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen3/qwen3_loader.py |
Add docstrings to existing functions | # pylint: disable=too-many-instance-attributes
from typing import Any, Dict, List, Literal, Optional, Union
from pydantic import BaseModel, Field
from mlc_llm.support.constants import MLC_CHAT_CONFIG_VERSION
from .conversation_protocol import Conversation
MLC_CHAT_SYSTEM_DEFAULT = {
"pad_token_id": 0,
"bos... | --- +++ @@ -1,4 +1,5 @@ # pylint: disable=too-many-instance-attributes
+"""Schema for mlc-chat-config"""
from typing import Any, Dict, List, Literal, Optional, Union
@@ -22,6 +23,7 @@
class MLCChatConfig(BaseModel):
+ """Fields in the dumped `mlc-chat-config.json` file."""
# Version control
vers... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/protocol/mlc_chat_config.py |
Add minimal docstrings for each function |
import tvm
from tvm.script import tir as T
from mlc_llm.support.max_thread_check import get_max_num_threads_per_block
# mypy: disable-error-code="attr-defined,valid-type,name-defined"
# pylint: disable=too-many-locals,invalid-name,too-many-arguments,unnecessary-lambda
# pylint: disable=too-many-statements,line-too-l... | --- +++ @@ -1,3 +1,4 @@+"""Operators for choosing the pivot to cut-off top-p percentile"""
import tvm
from tvm.script import tir as T
@@ -10,6 +11,32 @@
def top_p_pivot(pN, target: tvm.target.Target):
+ """Top-p pivot function. This function finds the pivot to cut-off top-p percentile.
+
+ A valide pivot ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/top_p_pivot.py |
Add docstrings following best practices |
import functools
from mlc_llm.loader import ExternMapping
from mlc_llm.quantization import Quantization
from .phi3v_model import Phi3VConfig, Phi3VForCausalLM
# pylint: disable=too-many-statements
def huggingface(model_config: Phi3VConfig, quantization: Quantization) -> ExternMapping:
model = Phi3VForCausalLM(... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's Phi parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -9,6 +13,22 @@
# pylint: disable=too-many-statements
def huggingface(model_config: Phi3VConfig, quantization: Quantization) -> Ext... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/phi3v/phi3v_loader.py |
Write docstrings describing functionality |
# pylint: disable=invalid-name
from typing import List, Literal, Tuple
import tvm
from tvm.relax.frontend import nn
from tvm.script import ir as I
from tvm.script import tir as T
try:
import triton
import triton.language as tl
except ImportError:
triton = None
tl = None
# We use a wrapper function... | --- +++ @@ -1,3 +1,4 @@+"""Operators enabled by external modules."""
# pylint: disable=invalid-name
@@ -52,6 +53,10 @@ BLOCK_SIZE_K: tl.constexpr,
GROUP_SIZE_M: tl.constexpr,
):
+ """Triton-accelerated function used to perform linear operations (dot
+ product) on input tensors `... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/triton.py |
Add docstrings to improve collaboration |
import dataclasses
from typing import Any, Dict, Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from mlc_llm.support import tensor_parall... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for Nemotron architecture.
+"""
import dataclasses
from typing import Any, Dict, Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class NemotronConfig(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Nemotron model."""
v... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/nemotron/nemotron_model.py |
Write beginner-friendly docstrings |
from enum import Enum
from typing import Any, Dict, List, Optional, Tuple, Type, TypeVar, Union
from pydantic import BaseModel, Field, field_validator
# The message placeholders in the message prompts according to roles.
class MessagePlaceholders(Enum):
SYSTEM = "{system_message}"
USER = "{user_message}"
... | --- +++ @@ -1,3 +1,4 @@+"""The standard conversation protocol in MLC LLM"""
from enum import Enum
from typing import Any, Dict, List, Optional, Tuple, Type, TypeVar, Union
@@ -7,6 +8,7 @@
# The message placeholders in the message prompts according to roles.
class MessagePlaceholders(Enum):
+ """The message pl... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/protocol/conversation_protocol.py |
Create docstrings for reusable components |
import dataclasses
from typing import Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.model.qwen2.qwen2_model import ACT2FN, QWen2Attention, QWen2Config
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for QWEN2MOE architecture.
+"""
import dataclasses
from typing import Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class Qwen2MoeConfig(QWen2Config): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Qwen2Moe model."""
moe_interme... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen2_moe/qwen2_moe_model.py |
Turn comments into proper docstrings |
from typing import List
from tvm import relax
from tvm.relax.frontend.nn import Tensor, op
def pipeline_stage_boundary(*tensors: Tensor) -> List[Tensor]:
# pylint: disable=protected-access
return op.wrap_nested(
relax.call_pure_packed(
"mlc.pipeline_parallel_stage_boundary",
... | --- +++ @@ -1,3 +1,4 @@+"""Operators for pipeline parallelism."""
from typing import List
@@ -6,6 +7,18 @@
def pipeline_stage_boundary(*tensors: Tensor) -> List[Tensor]:
+ """Pipeline parallelism stage boundary mark operator in MLC.
+
+ Parameters
+ ----------
+ tensors : Tensor
+ The tensor... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/pipeline_parallel.py |
Add documentation for all methods |
# pylint: disable=missing-class-docstring, disable=too-many-instance-attributes
from typing import Dict, List, Optional
from pydantic import BaseModel
from .debug_protocol import DebugConfig
from .openai_api_protocol import RequestResponseFormat
class GenerationConfig(BaseModel): # pylint:
n: int = 1
tem... | --- +++ @@ -1,3 +1,4 @@+"""Low-level generation config class"""
# pylint: disable=missing-class-docstring, disable=too-many-instance-attributes
from typing import Dict, List, Optional
@@ -9,6 +10,10 @@
class GenerationConfig(BaseModel): # pylint:
+ """The generation configuration dataclass.
+
+ This is a... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/protocol/generation_config.py |
Add docstrings explaining edge cases |
import functools
from dataclasses import dataclass
from typing import Any, Dict, List, Literal, Optional, Sequence, Tuple, Type, Union
import numpy as np
from tvm import DataType, DataTypeCode, IRModule, relax, runtime, te, tir, topi
from tvm.relax.frontend import nn
from tvm.runtime import Tensor
from mlc_llm.loade... | --- +++ @@ -1,3 +1,4 @@+"""The per-tensor quantization config"""
import functools
from dataclasses import dataclass
@@ -27,6 +28,7 @@
@dataclass
class PerTensorQuantize: # pylint: disable=too-many-instance-attributes
+ """Configuration for per-tensor quantization"""
name: str
kind: str
@@ -63,6 +... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/quantization/per_tensor_quantization.py |
Provide clean and structured docstrings |
from http import HTTPStatus
from typing import Optional
import fastapi
from pydantic import BaseModel
class BadRequestError(ValueError):
def __init__(self, *args: object) -> None:
super().__init__(*args)
class ErrorResponse(BaseModel):
object: str = "error"
message: str
code: Optional[in... | --- +++ @@ -1,3 +1,4 @@+"""Error protocols in MLC LLM"""
from http import HTTPStatus
from typing import Optional
@@ -7,12 +8,14 @@
class BadRequestError(ValueError):
+ """The exception for bad requests in engines."""
def __init__(self, *args: object) -> None:
super().__init__(*args)
clas... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/protocol/error_protocol.py |
Fill in missing docstrings in my code |
# pylint: disable=too-many-lines
import asyncio
import queue
import sys
import weakref
from typing import (
Any,
AsyncGenerator,
Dict,
Iterator,
List,
Literal,
Optional,
Tuple,
Union,
overload,
)
from tvm.runtime import Device
from mlc_llm.protocol import debug_protocol, open... | --- +++ @@ -1,3 +1,4 @@+"""The MLC LLM Serving Engine."""
# pylint: disable=too-many-lines
@@ -34,6 +35,7 @@
# Note: we define both AsyncChat and Chat for Python type analysis.
class AsyncChat: # pylint: disable=too-few-public-methods
+ """The proxy class to direct to async chat completions."""
def _... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/engine.py |
Add docstrings to clarify complex logic |
# pylint: disable=missing-class-docstring
import json
import time
from typing import Any, Dict, List, Literal, Optional, Tuple, Union
import shortuuid
from pydantic import BaseModel, Field, field_validator, model_validator
from .conversation_protocol import Conversation
from .debug_protocol import DebugConfig
from ... | --- +++ @@ -1,3 +1,7 @@+"""Protocols in MLC LLM for OpenAI API.
+Adapted from FastChat's OpenAI protocol:
+https://github.com/lm-sys/FastChat/blob/main/fastchat/protocol/openai_api_protocol.py
+"""
# pylint: disable=missing-class-docstring
@@ -68,6 +72,9 @@
class EmbeddingRequest(BaseModel):
+ """OpenAI "v1... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/protocol/openai_api_protocol.py |
Add docstrings including usage examples |
# pylint: disable=too-many-lines
import ast
import asyncio
import json
import numbers
import queue
import sys
import threading
from dataclasses import dataclass
from pathlib import Path
from typing import Any, Callable, Dict, List, Literal, Optional, Tuple, Union
import tvm
from tvm.runtime import Device
from mlc_l... | --- +++ @@ -1,3 +1,4 @@+"""The MLC LLM Serving engine base class."""
# pylint: disable=too-many-lines
@@ -32,6 +33,20 @@
@dataclass
class ModelInfo:
+ """The model info dataclass.
+
+ Parameters
+ ----------
+ model : str
+ The identifier of the input model.
+ It may be a compiled mode... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/engine_base.py |
Add documentation for all methods |
import dataclasses
from typing import Optional
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.model.qwen3.qwen3_model import ACT2FN, Qwen3Attention, Qwen3Config
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for QWEN2MOE architecture.
+"""
import dataclasses
from typing import Optional
@@ -18,6 +21,7 @@
@dataclasses.dataclass
class Qwen3MoeConfig(Qwen3Config): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Qwen3Moe model."""
moe_interme... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen3_moe/qwen3_moe_model.py |
Add minimal docstrings for each function |
import fastapi
from fastapi.responses import PlainTextResponse
from mlc_llm.serve.server import ServerContext
app = fastapi.APIRouter()
################ /metrics ################
@app.get("/metrics", response_class=PlainTextResponse)
async def metrics(_request: fastapi.Request):
server_context: ServerContext ... | --- +++ @@ -1,3 +1,4 @@+"""MLC LLM server metrics entrypoints"""
import fastapi
from fastapi.responses import PlainTextResponse
@@ -11,6 +12,7 @@
@app.get("/metrics", response_class=PlainTextResponse)
async def metrics(_request: fastapi.Request):
+ """Start the cuda profiler for the engine. Only for debug pur... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/entrypoints/metrics_entrypoints.py |
Provide clean and structured docstrings |
from dataclasses import dataclass, field
from typing import Any, Callable, Dict, List, Optional
from tvm import DataType, DataTypeCode, te, tir, topi
from tvm.relax.frontend import nn
from tvm.runtime import Tensor
from mlc_llm.loader import QuantizeMapping
from .utils import convert_uint_to_float, is_final_fc, is_... | --- +++ @@ -1,3 +1,4 @@+"""AWQ Quantization"""
from dataclasses import dataclass, field
from typing import Any, Callable, Dict, List, Optional
@@ -32,6 +33,7 @@
@dataclass
class AWQQuantize: # pylint: disable=too-many-instance-attributes
+ """Configuration for AWQ quantization"""
name: str
kind: ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/quantization/awq_quantization.py |
Expand my code with proper documentation strings |
import uuid
from typing import Any, Callable, Dict, List, Literal, Optional, Union
from mlc_llm.protocol import error_protocol, openai_api_protocol
from mlc_llm.protocol.generation_config import GenerationConfig
from mlc_llm.serve import data
RequestProtocol = Union[
openai_api_protocol.CompletionRequest, openai... | --- +++ @@ -1,3 +1,4 @@+"""Utility functions for MLC Serve engine"""
import uuid
from typing import Any, Callable, Dict, List, Literal, Optional, Union
@@ -12,6 +13,9 @@
def get_unsupported_fields(request: RequestProtocol) -> List[str]:
+ """Get the unsupported fields of the request.
+ Return the list of ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/engine_utils.py |
Write documentation strings for class attributes |
import json
from http import HTTPStatus
import fastapi
from mlc_llm.protocol import error_protocol
from mlc_llm.serve.server import ServerContext
app = fastapi.APIRouter()
################ /debug/dump_event_trace ################
@app.post("/debug/dump_event_trace")
async def debug_dump_event_trace(request: fast... | --- +++ @@ -1,3 +1,4 @@+"""MLC LLM server debug entrypoints"""
import json
from http import HTTPStatus
@@ -14,6 +15,10 @@
@app.post("/debug/dump_event_trace")
async def debug_dump_event_trace(request: fastapi.Request):
+ """Return the recorded events in Chrome Trace Event Format in JSON string.
+ The input... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/entrypoints/debug_entrypoints.py |
Write Python docstrings for this snippet |
from typing import Callable, List, Optional, Sequence
from tvm import IRModule, relax, te, tir
from tvm.relax.frontend import nn
from tvm.runtime import DataType, DataTypeCode
from tvm.s_tir import dlight as dl
from tvm.target import Target
from mlc_llm.support import tensor_parallel as tp
def convert_uint_to_floa... | --- +++ @@ -1,3 +1,4 @@+"""Common utilities for quantization"""
from typing import Callable, List, Optional, Sequence
@@ -20,6 +21,7 @@ out_shape: Optional[List[tir.PrimExpr]] = None,
ft_reorder: Optional[bool] = False,
) -> te.Tensor:
+ """Convert a quantized uint weight to an unquantized float weigh... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/quantization/utils.py |
Create docstrings for each class method |
from functools import reduce
from typing import Literal, Optional, Tuple, Union
import numpy as np
from tvm import te, tir
from tvm.relax.frontend.nn import IntExpr, Tensor, op
from tvm.script import tir as T
# mypy: disable-error-code="attr-defined,name-defined"
# pylint: disable=line-too-long,too-many-locals,inval... | --- +++ @@ -1,3 +1,4 @@+"""Mixture of Experts operators"""
from functools import reduce
from typing import Literal, Optional, Tuple, Union
@@ -12,6 +13,9 @@
def moe_sum(x: Tensor, dim: int) -> Tensor:
+ """Compute the sum of the input tensor along the given axis. It is specialized for the MoE
+ case where... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/moe_misc.py |
Generate docstrings for this script |
import json
import math
import threading
from typing import Any, AsyncGenerator, Iterable, List, Literal, Optional, Tuple
import aiohttp # pylint: disable=import-error
import tvm
from mlc_llm.protocol import openai_api_protocol
from mlc_llm.serve import EngineConfig, PopenServer
from mlc_llm.serve.entrypoints impor... | --- +++ @@ -1,3 +1,4 @@+"""Programmable router for dispatching OpenAI API to Microserving API"""
import json
import math
@@ -14,6 +15,7 @@
class Router: # pylint: disable=too-many-instance-attributes
+ """Programmable Router Implementation"""
def __init__(
self,
@@ -26,6 +28,9 @@ rou... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/router/router.py |
Add docstrings to clarify complex logic |
from dataclasses import dataclass
from typing import Dict, List, Optional, Tuple
import tvm
import tvm_ffi
from tvm.runtime import Object, Tensor
from . import _ffi_api
@tvm_ffi.register_object("mlc.serve.Data") # pylint: disable=protected-access
class Data(Object): # pylint: disable=too-few-public-methods
... | --- +++ @@ -1,3 +1,4 @@+"""Classes denoting multi-modality data used in MLC LLM serving"""
from dataclasses import dataclass
from typing import Dict, List, Optional, Tuple
@@ -11,6 +12,7 @@
@tvm_ffi.register_object("mlc.serve.Data") # pylint: disable=protected-access
class Data(Object): # pylint: disable=too-f... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/data.py |
Generate documentation strings for clarity |
# pylint: disable=too-many-locals,too-many-return-statements,too-many-statements,fixme
import base64
import struct
from http import HTTPStatus
from typing import AsyncGenerator, List, Optional
import fastapi
import numpy as np
from mlc_llm.protocol import error_protocol
from mlc_llm.protocol.openai_api_protocol impo... | --- +++ @@ -1,3 +1,4 @@+"""OpenAI API-compatible server entrypoints in MLC LLM"""
# pylint: disable=too-many-locals,too-many-return-statements,too-many-statements,fixme
import base64
@@ -26,6 +27,7 @@
def verify_api_key(request: fastapi.Request):
+ """Function to verify API key"""
server_context = Serve... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/entrypoints/openai_entrypoints.py |
Add docstrings that explain logic |
import functools
import numpy as np
from ...loader import ExternMapping
from ...quantization import Quantization
from .rwkv5_model import RWKV5_ForCausalLM, RWKV5Config
def huggingface(model_config: RWKV5Config, quantization: Quantization) -> ExternMapping:
model = RWKV5_ForCausalLM(model_config)
if quanti... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's RWKV5 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
@@ -9,6 +13,22 @@
def huggingface(model_config: RWKV5Config, quantization: Quantization) -> ExternMapping:
+ """Returns a param... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/rwkv5/rwkv5_loader.py |
Add docstrings including usage examples |
import json
from dataclasses import asdict, dataclass, field
from typing import List, Literal, Optional, Tuple, Union
@dataclass
class EngineConfig: # pylint: disable=too-many-instance-attributes
model: Optional[str] = None
model_lib: Optional[str] = None
additional_models: List[Union[str, Tuple[str, s... | --- +++ @@ -1,3 +1,4 @@+"""Configuration dataclasses used in MLC LLM serving"""
import json
from dataclasses import asdict, dataclass, field
@@ -6,6 +7,132 @@
@dataclass
class EngineConfig: # pylint: disable=too-many-instance-attributes
+ """The class of MLCEngine execution configuration.
+
+ Parameters
+... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/config.py |
Generate helpful docstrings for debugging |
from tvm.target import Target
def get_max_num_threads_per_block(target: Target) -> int:
max_num_threads = target.attrs.get("max_num_threads")
max_threads_per_block = target.attrs.get("max_threads_per_block", None)
if max_threads_per_block is None:
return max_num_threads
return max(max_num_thr... | --- +++ @@ -1,8 +1,13 @@+"""Helper functions for checking max num thread."""
from tvm.target import Target
def get_max_num_threads_per_block(target: Target) -> int:
+ """
+ max(max_num_threads, max_threads_per_block); if latter does not exist, return max_num_threads.
+ We add this method since some tar... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/max_thread_check.py |
Help me comply with documentation standards |
import logging
from typing import Any, Callable, Dict, Sequence, Tuple
from tvm import IRModule, relax
from tvm.relax.frontend import nn
from tvm.runtime import Device, Tensor
from tvm.s_tir import dlight as dl
from tvm.target import Target
logger = logging.getLogger("preshard")
def _sharded_param_name(param_name,... | --- +++ @@ -1,3 +1,4 @@+"""Functions for pre-sharding weights"""
import logging
from typing import Any, Callable, Dict, Sequence, Tuple
@@ -72,6 +73,23 @@ tensor_parallel_shards: int,
args: Any,
) -> Tuple[Dict[str, nn.Parameter], Dict[str, Callable[[Tensor], Sequence[Tensor]]]]:
+ """Apply pre-shardin... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/preshard.py |
Add docstrings for utility scripts |
import asyncio
import concurrent.futures
import json
import os
from typing import List, Literal, Optional, Tuple, Union
import numpy as np
import tvm
from tvm import relax
from tvm.runtime import Device, ShapeTuple
from mlc_llm.serve import engine_utils
from mlc_llm.support.auto_device import detect_device
from mlc_... | --- +++ @@ -1,3 +1,4 @@+"""Asynchronous embedding inference engine for encoder and decoder models."""
import asyncio
import concurrent.futures
@@ -16,6 +17,28 @@
class AsyncEmbeddingEngine: # pylint: disable=too-many-instance-attributes
+ """Asynchronous embedding inference engine.
+
+ Supports both enco... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/embedding_engine.py |
Document all public functions with docstrings |
# pylint: disable=too-many-statements,too-many-lines,too-many-arguments
import json
from typing import Any, Dict, List, Literal, Optional, Union
import numpy as np
from tvm import relax as rx
from tvm import tir
from tvm.relax.frontend.nn.llm.kv_cache import PagedKVCache as TVMPagedKVCache
from tvm.relax.frontend.nn.... | --- +++ @@ -1,3 +1,4 @@+"""Attention KV cache modeling."""
# pylint: disable=too-many-statements,too-many-lines,too-many-arguments
import json
@@ -11,6 +12,7 @@
class PagedKVCache(TVMPagedKVCache): # pylint: disable=too-few-public-methods
+ """The Paged KV Cache used in LLM batching for efficient attention ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/nn/kv_cache.py |
Add docstrings to existing functions |
import functools
from typing import Callable, List
import numpy as np
from mlc_llm.loader import ExternMapping, QuantizeMapping
from mlc_llm.quantization import BlockScaleQuantize, Quantization
from .qwen3_moe_model import Qwen3MoeConfig, Qwen3MoeForCausalLM
def huggingface(model_config: Qwen3MoeConfig, quantizat... | --- +++ @@ -1,3 +1,7 @@+"""
+This file specifies how MLC's QWen2 parameter maps from other formats, for example HuggingFace
+PyTorch, HuggingFace safetensors.
+"""
import functools
from typing import Callable, List
@@ -11,6 +15,22 @@
def huggingface(model_config: Qwen3MoeConfig, quantization: Quantization) -> E... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen3_moe/qwen3_moe_loader.py |
Document all public functions with docstrings |
import dataclasses
from functools import partial
from typing import Any, Dict, Optional, Tuple
from tvm import te, tir
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import Tensor, op
from mlc_llm import op as op_ext
from mlc_llm.nn import PagedKVCache, RopeMode
from mlc_llm.support import logging
from... | --- +++ @@ -1,3 +1,6 @@+"""
+Implementation for QWEN2 architecture.
+"""
import dataclasses
from functools import partial
@@ -19,6 +22,7 @@
@dataclasses.dataclass
class Qwen3Config(ConfigBase): # pylint: disable=too-many-instance-attributes
+ """Configuration of the Qwen3 model."""
hidden_act: str
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/model/qwen3/qwen3_model.py |
Write docstrings including parameters and return values |
from dataclasses import dataclass
from typing import Any, Literal, Optional, Tuple
import tvm
from tvm import DataType, DataTypeCode, te, tir
from tvm.relax.frontend import nn
from tvm.script import tir as T
from mlc_llm.loader import QuantizeMapping
from mlc_llm.nn import MixtralExperts
from mlc_llm.op import cutla... | --- +++ @@ -1,3 +1,4 @@+"""The block-scale quantization config"""
from dataclasses import dataclass
from typing import Any, Literal, Optional, Tuple
@@ -20,6 +21,7 @@
@dataclass
class BlockScaleQuantize: # pylint: disable=too-many-instance-attributes
+ """Configuration for block-scale quantization"""
... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/quantization/block_scale_quantization.py |
Write docstrings describing each step |
import numpy as np
from tvm import relax, runtime
from tvm.relax.frontend import nn
from mlc_llm.nn import MixtralExperts
from ..op import cutlass, extern, moe_matmul
from . import per_tensor_quantization as ptq
from .utils import apply_sharding
class FP8PerTensorQuantizeMixtralExperts(
ptq.PerTensorQuantizeMi... | --- +++ @@ -1,3 +1,4 @@+"""Quantization techniques for FP8"""
import numpy as np
from tvm import relax, runtime
@@ -13,6 +14,7 @@ class FP8PerTensorQuantizeMixtralExperts(
ptq.PerTensorQuantizeMixtralExperts
): # pylint: disable=too-many-instance-attributes
+ """MixtralExperts with per-tensor quantization... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/quantization/fp8_quantization.py |
Add docstrings to improve collaboration |
import logging
import os
def enable_logging():
if os.getenv("MLC_UNSET_LOGGING"):
return
logging.basicConfig(
level=logging.INFO,
style="{",
datefmt="%Y-%m-%d %H:%M:%S",
format="[{asctime}] {levelname} {filename}:{lineno}: {message}",
)
def getLogger(name: str): ... | --- +++ @@ -1,9 +1,14 @@+"""
+Logging support for MLC. It derives from Python's logging module, and in the future,
+it can be easily replaced by other logging modules such as structlog.
+"""
import logging
import os
def enable_logging():
+ """Enable MLC's default logging format"""
if os.getenv("MLC_UNS... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/logging.py |
Write beginner-friendly docstrings |
import fastapi
from mlc_llm.protocol.debug_protocol import DisaggConfig
from mlc_llm.protocol.microserving_protocol import (
PrepRecvRequest,
PrepRecvResponse,
RemoteSendRequest,
StartGenerateRequest,
)
from mlc_llm.protocol.openai_api_protocol import StreamOptions
from .openai_entrypoints import req... | --- +++ @@ -1,3 +1,4 @@+"""MicroServing server entrypoints in MLC LLM"""
import fastapi
@@ -20,6 +21,11 @@
@app.post("/microserving/prep_recv")
async def prep_recv(request: PrepRecvRequest, raw_request: fastapi.Request) -> PrepRecvResponse:
+ """Handle the microserving request for receive preparation.
+ M... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/entrypoints/microserving_entrypoints.py |
Add docstrings that explain inputs and outputs |
import sys
def set_global_random_seed(seed):
if "numpy" in sys.modules:
sys.modules["numpy"].random.seed(seed)
if "torch" in sys.modules:
sys.modules["torch"].manual_seed(seed)
if "random" in sys.modules:
sys.modules["random"].seed(seed) # pylint: disable=no-member
if "tvm" i... | --- +++ @@ -1,8 +1,10 @@+"""Utility functions for random number generation."""
import sys
def set_global_random_seed(seed):
+ """Set global random seed for python, numpy, torch and tvm."""
if "numpy" in sys.modules:
sys.modules["numpy"].random.seed(seed)
if "torch" in sys.modules:
@@ -12,4... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/random.py |
Document functions with detailed explanations |
from typing import Optional, Tuple
from tvm.relax.frontend import nn
from tvm.relax.frontend.nn import op
def group_gemm(
x: nn.Tensor,
weight: nn.Tensor,
indptr: nn.Tensor,
scale: Optional[nn.Tensor] = None,
weight_dtype: Optional[str] = None,
out_dtype: Optional[str] = None,
): # pylint: ... | --- +++ @@ -1,3 +1,4 @@+"""Operators enabled by external modules."""
from typing import Optional, Tuple
@@ -13,6 +14,34 @@ weight_dtype: Optional[str] = None,
out_dtype: Optional[str] = None,
): # pylint: disable=too-many-arguments
+ """
+ Cutlass group gemm operator.
+
+ Parameters
+ ------... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/op/cutlass.py |
Write docstrings for utility functions |
from typing import Any, Callable, Dict, Optional, Tuple, Type
from tvm.relax.frontend import nn
from mlc_llm.loader import QuantizeMapping
from .awq_quantization import AWQQuantize
from .block_scale_quantization import BlockScaleQuantize
from .ft_quantization import FTQuantize
from .group_quantization import GroupQ... | --- +++ @@ -1,3 +1,4 @@+"""Quantization factory utilities for model quantization."""
from typing import Any, Callable, Dict, Optional, Tuple, Type
@@ -29,6 +30,7 @@ set_tensor_parallel_shards: bool = True,
per_tensor_use_shards: bool = True,
) -> Dict[str, FuncQuantization]:
+ """Create standard quant... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/quantization/model_quantization.py |
Fill in missing docstrings in my code |
import concurrent.futures as cf
import hashlib
import json
import os
import shutil
import subprocess
import tempfile
from pathlib import Path
from typing import List, Optional, Tuple
import requests # pylint: disable=import-error
from . import logging, tqdm
from .constants import (
MLC_DOWNLOAD_CACHE_POLICY,
... | --- +++ @@ -1,3 +1,4 @@+"""Common utilities for downloading files from HuggingFace or other URLs online."""
import concurrent.futures as cf
import hashlib
@@ -24,6 +25,7 @@
def log_download_cache_policy():
+ """log current download policy"""
logger.info(
"%s = %s. Can be one of: ON, OFF, REDO, ... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/download_cache.py |
Generate docstrings with examples |
from dataclasses import dataclass
from functools import partial
from typing import Any, List, Literal, Optional, Tuple, Union
from tvm import DataType, DataTypeCode, IRModule, relax, te, tir, topi
from tvm.relax.frontend import nn
from tvm.runtime import Tensor
from mlc_llm.loader import QuantizeMapping
from mlc_llm... | --- +++ @@ -1,3 +1,4 @@+"""The group quantization config"""
from dataclasses import dataclass
from functools import partial
@@ -25,6 +26,7 @@
@dataclass
class GroupQuantize: # pylint: disable=too-many-instance-attributes
+ """Configuration for group quantization"""
name: str
kind: str
@@ -66,6 +6... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/quantization/group_quantization.py |
Document all endpoints with docstrings |
import json
from typing import Any, Callable, Dict, List, Literal, Optional, Sequence, Tuple, Union
import tvm
from mlc_llm.protocol.generation_config import GenerationConfig
from mlc_llm.serve import data
from mlc_llm.serve.config import EngineConfig
from mlc_llm.serve.engine_base import (
EngineMetrics,
_c... | --- +++ @@ -1,3 +1,12 @@+"""The MLC LLM synchronized engine.
+
+NOTE: This engine defined in this file directly wraps the underlying
+Engine implementation in C++, is not optimized by multi-threading and
+does not offer standard OpenAI API interface.
+
+We do not expose it and use it by default. As of now it mainly ser... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/sync_engine.py |
Help me write clear docstrings |
import dataclasses
from contextlib import contextmanager
from typing import Any, Dict, List, Optional
from tvm import te, tir, topi
from tvm.relax.frontend import nn
@dataclasses.dataclass
class ShardSingleDim:
name: str
dim: int
segs: Optional[List[int]] = None
def gen_tir(self, shards: int, weig... | --- +++ @@ -1,3 +1,4 @@+"""Sharding operators for tensor parallelism."""
import dataclasses
from contextlib import contextmanager
@@ -9,12 +10,31 @@
@dataclasses.dataclass
class ShardSingleDim:
+ """
+ Shard a tensor by a single dimension.
+
+
+ Parameters
+ ----------
+ name : str
+ The na... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/tensor_parallel.py |
Add docstrings for internal functions |
# pylint: disable=import-error
# isort: off
import json
import os
from typing import Dict, List, Optional
def bpe(
mergeable_ranks: Dict[bytes, int], token: bytes, max_rank: Optional[int] = None
) -> List[bytes]:
parts = [bytes([b]) for b in token]
while True:
min_idx = None
min_rank = No... | --- +++ @@ -1,3 +1,7 @@+"""
+Adapted from https://gist.github.com/xenova/a452a6474428de0182b17605a98631ee
+Generator of mlc-chat-config.json and tokenizer configuration.
+"""
# pylint: disable=import-error
# isort: off
@@ -9,6 +13,7 @@ def bpe(
mergeable_ranks: Dict[bytes, int], token: bytes, max_rank: Optiona... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/convert_tiktoken.py |
Create Google-style docstrings for my code |
import os
from pathlib import Path
from typing import TYPE_CHECKING, Callable, List, Optional, Tuple
from tvm import IRModule, relax
from tvm.contrib import ndk, tar, xcode
from tvm.ir.transform import Pass
from tvm.target import Target
from tvm_ffi import get_global_func, register_global_func
from . import logging
... | --- +++ @@ -1,3 +1,4 @@+"""Helper functions for target auto-detection."""
import os
from pathlib import Path
@@ -28,6 +29,17 @@
def detect_target_and_host(target_hint: str, host_hint: str = "auto") -> Tuple[Target, BuildFunc]:
+ """Detect the configuration for the target device and its host, for example, tar... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/auto_target.py |
Add return value explanations in docstrings |
import json
import tempfile
from pathlib import Path
from typing import TYPE_CHECKING
from . import logging
from .style import bold, green
if TYPE_CHECKING:
from mlc_llm.model import Model # pylint: disable=unused-import
from mlc_llm.quantization import Quantization # pylint: disable=unused-import
logger... | --- +++ @@ -1,3 +1,4 @@+"""Help function for detecting the model configuration file `config.json`"""
import json
import tempfile
@@ -18,6 +19,20 @@
def detect_mlc_chat_config(mlc_chat_config: str) -> Path:
+ """Detect and return the path that points to mlc-chat-config.json.
+ If `mlc_chat_config` is a dir... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/auto_config.py |
Create structured documentation for my script |
from typing import TYPE_CHECKING, Dict, List, Optional
from ..engine import AsyncMLCEngine
if TYPE_CHECKING:
from ..embedding_engine import AsyncEmbeddingEngine
class ServerContext:
server_context: Optional["ServerContext"] = None
enable_debug: bool = False
def __init__(self) -> None:
sel... | --- +++ @@ -1,3 +1,4 @@+"""Server context that shared by multiple entrypoint files."""
from typing import TYPE_CHECKING, Dict, List, Optional
@@ -8,6 +9,9 @@
class ServerContext:
+ """The global server context, including the running models
+ and corresponding async engines.
+ """
server_context... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/serve/server/server_context.py |
Write docstrings that follow conventions | import sqlite3
import uuid
import json
from datetime import datetime
from typing import List, Dict, Optional, Tuple
class ChatDatabase:
def __init__(self, db_path: str = None):
if db_path is None:
# Auto-detect environment and set appropriate path
import os
if os.path.ex... | --- +++ @@ -18,6 +18,7 @@ self.init_database()
def init_database(self):
+ """Initialize the SQLite database with required tables"""
conn = sqlite3.connect(self.db_path)
cursor = conn.cursor()
@@ -105,6 +106,7 @@ print("✅ Database initialized successfully")
... | https://raw.githubusercontent.com/PromtEngineer/localGPT/HEAD/backend/database.py |
Write docstrings that follow conventions | # pylint: disable=missing-docstring
import argparse
import logging
import os
import re
import subprocess
# Modify the following value during release
# ---------------------------------------------------
# Current version:
# We use the version of the incoming release for code
# that is under development.
#
# It is also... | --- +++ @@ -1,4 +1,11 @@ # pylint: disable=missing-docstring
+"""
+This is the global script that set the version information of TVM.
+This script runs and update all the locations that related to versions
+
+List of affected files:
+- mlc-llm-root/pyproject.toml
+"""
import argparse
import logging
import os
@@ -30,... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/version.py |
Generate missing documentation strings |
import os
import subprocess
import sys
from typing import Dict, Optional
import tvm
from tvm.runtime import Device
from tvm_ffi import DLDeviceType
from . import logging
from .style import bold, green, red
FOUND = green("Found")
NOT_FOUND = red("Not found")
AUTO_DETECT_DEVICES = ["cuda", "rocm", "metal", "vulkan", ... | --- +++ @@ -1,3 +1,4 @@+"""Automatic detection of the device available on the local machine."""
import os
import subprocess
@@ -21,6 +22,7 @@
def detect_device(device_hint: str) -> Optional[Device]:
+ """Detect locally available device from string hint."""
if device_hint == "auto":
device = Non... | https://raw.githubusercontent.com/mlc-ai/mlc-llm/HEAD/python/mlc_llm/support/auto_device.py |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.