Optimize the log of the entropy coeff instead of the entropy coeff (#56)

* optimize the log of the entropy coeff instead of the entropy coeff * Update log ent coef for SAC and derivates * Reformat yaml * Use uv for faster downloads * Remove TODO * Remove redundant call --------- Co-authored-by: Antonin RAFFIN <[email protected]>
araffin · Nov 1, 2024 · 1c79684 · 1c79684
1 parent 19c85a1
commit 1c79684
Show file tree

Hide file tree

Showing 6 changed files with 45 additions and 40 deletions.
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -5,9 +5,9 @@ name: CI
 
 on:
   push:
-    branches: [ master ]
+    branches: [master]
   pull_request:
-    branches: [ master ]
+    branches: [master]
 
 jobs:
   build:
@@ -23,34 +23,37 @@ jobs:
         python-version: ["3.8", "3.9", "3.10", "3.11"]
 
     steps:
-    - uses: actions/checkout@v3
-    - name: Set up Python ${{ matrix.python-version }}
-      uses: actions/setup-python@v4
-      with:
-        python-version: ${{ matrix.python-version }}
-    - name: Install dependencies
-      run: |
-        python -m pip install --upgrade pip
-        # cpu version of pytorch
-        pip install torch==2.1.0 --index-url https://download.pytorch.org/whl/cpu
+      - uses: actions/checkout@v3
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v4
+        with:
+          python-version: ${{ matrix.python-version }}
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          # Use uv for faster downloads
+          pip install uv
+          # cpu version of pytorch
+          # See https://github.com/astral-sh/uv/issues/1497
+          uv pip install --system torch==2.3.1+cpu --index https://download.pytorch.org/whl/cpu
 
-        pip install .[tests]
-        # Use headless version
-        pip install opencv-python-headless
-    - name: Lint with ruff
-      run: |
-        make lint
-    # - name: Build the doc
-    #   run: |
-    #     make doc
-    - name: Check codestyle
-      run: |
-        make check-codestyle
-    - name: Type check
-      run: |
-        make type
-      # skip mypy, jax doesn't have its latest version for python 3.8
-      if: "!(matrix.python-version == '3.8')"
-    - name: Test with pytest
-      run: |
-        make pytest
+          uv pip install --system .[tests]
+          # Use headless version
+          uv pip install --system opencv-python-headless
+      - name: Lint with ruff
+        run: |
+          make lint
+      # - name: Build the doc
+      #   run: |
+      #     make doc
+      - name: Check codestyle
+        run: |
+          make check-codestyle
+      - name: Type check
+        run: |
+          make type
+        # skip mypy, jax doesn't have its latest version for python 3.8
+        if: "!(matrix.python-version == '3.8')"
+      - name: Test with pytest
+        run: |
+          make pytest
diff --git a/sbx/crossq/crossq.py b/sbx/crossq/crossq.py
@@ -355,8 +355,10 @@ def actor_loss(
     @jax.jit
     def update_temperature(target_entropy: ArrayLike, ent_coef_state: TrainState, entropy: float):
         def temperature_loss(temp_params: flax.core.FrozenDict) -> jax.Array:
+            # Note: we optimize the log of the entropy coeff which is slightly different from the paper
+            # as discussed in https://github.com/rail-berkeley/softlearning/issues/37
             ent_coef_value = ent_coef_state.apply_fn({"params": temp_params})
-            ent_coef_loss = ent_coef_value * (entropy - target_entropy).mean()  # type: ignore[union-attr]
+            ent_coef_loss = jnp.log(ent_coef_value) * (entropy - target_entropy).mean()  # type: ignore[union-attr]
             return ent_coef_loss
 
         ent_coef_loss, grads = jax.value_and_grad(temperature_loss)(ent_coef_state.params)

diff --git a/sbx/dqn/policies.py b/sbx/dqn/policies.py
@@ -107,7 +107,6 @@ def build(self, key: jax.Array, lr_schedule: Schedule) -> jax.Array:
             ),
         )
 
-        # TODO: jit qf.apply_fn too?
         self.qf.apply = jax.jit(self.qf.apply)  # type: ignore[method-assign]
 
         return key

diff --git a/sbx/sac/sac.py b/sbx/sac/sac.py
@@ -141,8 +141,6 @@ def _setup_model(self) -> None:
                     ent_coef_init = float(self.ent_coef_init.split("_")[1])
                     assert ent_coef_init > 0.0, "The initial value of ent_coef must be greater than 0"
 
-                # Note: we optimize the log of the entropy coeff which is slightly different from the paper
-                # as discussed in https://github.com/rail-berkeley/softlearning/issues/37
                 self.ent_coef = EntropyCoef(ent_coef_init)
             else:
                 # This will throw an error if a malformed string (different from 'auto') is passed
@@ -325,8 +323,10 @@ def soft_update(tau: float, qf_state: RLTrainState) -> RLTrainState:
     @jax.jit
     def update_temperature(target_entropy: ArrayLike, ent_coef_state: TrainState, entropy: float):
         def temperature_loss(temp_params: flax.core.FrozenDict) -> jax.Array:
+            # Note: we optimize the log of the entropy coeff which is slightly different from the paper
+            # as discussed in https://github.com/rail-berkeley/softlearning/issues/37
             ent_coef_value = ent_coef_state.apply_fn({"params": temp_params})
-            ent_coef_loss = ent_coef_value * (entropy - target_entropy).mean()  # type: ignore[union-attr]
+            ent_coef_loss = jnp.log(ent_coef_value) * (entropy - target_entropy).mean()  # type: ignore[union-attr]
             return ent_coef_loss
 
         ent_coef_loss, grads = jax.value_and_grad(temperature_loss)(ent_coef_state.params)

diff --git a/sbx/tqc/tqc.py b/sbx/tqc/tqc.py
@@ -383,9 +383,10 @@ def soft_update(tau: float, qf1_state: RLTrainState, qf2_state: RLTrainState) ->
     @jax.jit
     def update_temperature(target_entropy: ArrayLike, ent_coef_state: TrainState, entropy: float):
         def temperature_loss(temp_params: flax.core.FrozenDict) -> jax.Array:
+            # Note: we optimize the log of the entropy coeff which is slightly different from the paper
+            # as discussed in https://github.com/rail-berkeley/softlearning/issues/37
             ent_coef_value = ent_coef_state.apply_fn({"params": temp_params})
-            # ent_coef_loss = (jnp.log(ent_coef_value) * (entropy - target_entropy)).mean()
-            ent_coef_loss = ent_coef_value * (entropy - target_entropy).mean()  # type: ignore[union-attr]
+            ent_coef_loss = jnp.log(ent_coef_value) * (entropy - target_entropy).mean()  # type: ignore[union-attr]
             return ent_coef_loss
 
         ent_coef_loss, grads = jax.value_and_grad(temperature_loss)(ent_coef_state.params)

diff --git a/sbx/version.txt b/sbx/version.txt
@@ -1 +1 @@
-0.17.0
+0.18.0