Upload CustomMixtralForCausalLM

README.md

@@ -0,0 +1,199 @@
+---
+library_name: transformers
+tags: []
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]

config.json

@@ -0,0 +1,33 @@
+{
+  "_name_or_path": "None",
+  "architectures": [
+    "CustomMixtralForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "auto_map": {
+    "AutoModelForCausalLM": "custom_mixtral.CustomMixtralForCausalLM"
+  },
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 2400,
+  "max_position_embeddings": 131072,
+  "model_type": "mixtral",
+  "num_attention_heads": 16,
+  "num_experts_per_tok": 1,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 8,
+  "num_local_experts": 2,
+  "output_router_logits": true,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 10000.0,
+  "router_aux_loss_coef": 0.01,
+  "sliding_window": 1024,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.39.1",
+  "use_cache": true,
+  "vocab_size": 32000
+}

custom_mixtral.py

@@ -0,0 +1,131 @@
+from transformers import MixtralForCausalLM, MixtralConfig
+from transformers.modeling_outputs import MoeCausalLMOutputWithPast, MoECausalLMOutputWithPast
+from transformers.models.mixtral.modeling_mixtral import MixtralDecoderLayer, MixtralSparseMoeBlock, load_balancing_loss_func
+from .noisy_gate import NoisyGate
+import torch
+import torch.nn as nn
+from typing import List, Optional, Tuple, Union
+
+def router_z_loss_func(
+    gate_logits: torch.Tensor, num_experts: torch.Tensor = None, top_k=2
+) -> float:
+    """Router z-loss used in ST-MoE."""
+    if gate_logits is None or not isinstance(gate_logits, tuple):
+        return 0
+
+    if isinstance(gate_logits, tuple):
+        compute_device = gate_logits[0].device
+        concatenated_gate_logits = torch.cat([layer_gate.to(compute_device) for layer_gate in gate_logits], dim=0)
+
+    router_z_loss = torch.logsumexp(concatenated_gate_logits, dim = -1)
+    router_z_loss = torch.square(router_z_loss)            
+    router_z_loss = router_z_loss.mean()
+
+    return router_z_loss
+
+class CustomMixtralConfig(MixtralConfig):
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+
+class CustomMixtralForCausalLM(MixtralForCausalLM):
+    """Mixtral with z-loss. Gating improvement based on ST-MoE."""
+    def __init__(self, config):
+        super().__init__(config)
+        self.router_z_loss_coef = 1e-3
+        for layer in self.model.layers:
+            layer.block_sparse_moe.gate = NoisyGate(config.hidden_size, config.num_local_experts, noise_mult=1.0, bias=False)
+        
+
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[List[torch.FloatTensor]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        output_router_logits: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+    ):
+        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
+        output_router_logits = (
+            output_router_logits if output_router_logits is not None else self.config.output_router_logits
+        )
+
+        output_hidden_states = (
+            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+        )
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+
+        # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            output_router_logits=output_router_logits,
+            return_dict=return_dict,
+        )
+
+        hidden_states = outputs[0]
+        logits = self.lm_head(hidden_states)
+        logits = logits.float()
+
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            shift_logits = logits[..., :-1, :].contiguous()
+            shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = nn.CrossEntropyLoss()
+            shift_logits = shift_logits.view(-1, self.config.vocab_size)
+            shift_labels = shift_labels.view(-1)
+            # Enable model parallelism
+            shift_labels = shift_labels.to(shift_logits.device)
+            loss = loss_fct(shift_logits, shift_labels)
+
+        aux_loss = None
+        if output_router_logits:
+            aux_loss = load_balancing_loss_func(
+                outputs.router_logits if return_dict else outputs[-1],
+                self.num_experts,
+                self.num_experts_per_tok,
+                attention_mask,
+            )
+            if labels is not None:
+                loss += self.router_aux_loss_coef * aux_loss.to(loss.device)  # make sure to reside in the same device
+
+        router_z_loss = None
+        if output_router_logits:
+            router_z_loss = router_z_loss_func(
+                outputs.router_logits if return_dict else outputs[-1],
+                self.num_experts,
+                self.num_experts_per_tok,
+            )
+            if labels is not None:
+                loss += self.router_z_loss_coef * router_z_loss.to(loss.device)
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            if output_router_logits:
+                output = (router_z_loss,) + output
+                output = (aux_loss,) + output
+            return (loss,) + output if loss is not None else output
+
+        return MoECausalLMOutputWithPast(
+            loss=loss,
+            aux_loss=aux_loss,
+            z_loss=router_z_loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+            router_logits=outputs.router_logits,
+        )

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "transformers_version": "4.39.1"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:80f0da679083496945248327ff3da30b282a5f119f83812614efa17f14be6214
+size 990092384

noisy_gate.py

@@ -0,0 +1,25 @@
+import torch
+import torch.nn as nn
+
+def log(t, eps = 1e-20):
+    return torch.log(t.clamp(min = eps))
+
+def gumbel_noise(t):
+    noise = torch.zeros_like(t).uniform_(0, 1)
+    return -log(-log(noise))
+
+class NoisyGate(nn.Module):
+    def __init__(self, hidden_dim, num_experts, noise_mult=1.0, bias=False):
+        super().__init__()
+        self.hidden_dim = hidden_dim
+        self.num_experts = num_experts
+        self.noise_mult = noise_mult
+        self.bias = bias
+        self.gate = nn.Linear(self.hidden_dim, self.num_experts, bias=self.bias)
+    
+    def forward(self, x):
+        x = self.gate(x)
+        noise = gumbel_noise(x)
+        out = x + noise * self.noise_mult
+        return out
+