docs: bump version references to 0.7 for the 0.7.0 release

Point the install constraints in the README and the eight example
livebooks at `{:emily, "~> 0.7"}`, and label the emily-vs-EXLA
benchmark reports as the 0.7.0 build.

The mix.exs @Version stays 0.6.1 — `mix publisho` bumps and tags it at
release time. Qwen3-0.6B (the model name) and the benchmark figures
are deliberately untouched.

Author: ausimian

README.md

@@ -29,7 +29,7 @@ Add `:emily` to your `mix.exs` deps:
 ```elixir
 def deps do
   [
-    {:emily, "~> 0.6"}
+    {:emily, "~> 0.7"}
   ]
 end
 ```
@@ -99,7 +99,7 @@ Releases on first `mix compile`.
 
 ### As a hex consumer
 
-Add `{:emily, "~> 0.6"}` to `mix.exs`, then:
+Add `{:emily, "~> 0.7"}` to `mix.exs`, then:
 
 ```sh
 mix deps.get

bench/emily_vs_exla_report.md

@@ -43,7 +43,7 @@ Emily's losses below are coverage gaps; they are all kernel/dispatch efficiency.
 | Date | 2026-06-07 |
 | Host | Apple M4 Pro, 12 cores, 24GB memory |
 | Elixir / OTP | 1.19.5 / 28 |
-| Emily | 0.6.1 (local checkout, AOT MLX variant) |
+| Emily | 0.7.0 (local checkout, AOT MLX variant) |
 | EXLA / XLA | 0.12.0 / 0.10.0 (host / CPU client) |
 | Nx | 0.12.1 |
 

bench/emily_vs_exla_results.md

@@ -10,7 +10,7 @@ higher is better for tokens/sec. Ratios are relative to the EXLA lane._
 | date | 2026-06-07 10:09 |
 | host | Apple M4 Pro (BEAM total 112 MB at write) |
 | elixir / otp | 1.19.5 / 28 |
-| emily | 0.6.1 |
+| emily | 0.7.0 |
 | exla | 0.12.0 (host (1 device)) |
 | nx | 0.12.1 |
 | smoke run? | false |

livebooks/distilbert_qa.livemd

@@ -3,7 +3,7 @@
 ```elixir
 Mix.install(
   [
-    {:emily, "~> 0.6"},
+    {:emily, "~> 0.7"},
     {:bumblebee, "~> 0.7"},
     {:tokenizers, "~> 0.5"},
     {:nx, "~> 0.12"},

livebooks/fast_kernels.livemd

@@ -3,7 +3,7 @@
 ```elixir
 Mix.install(
   [
-    {:emily, "~> 0.6"},
+    {:emily, "~> 0.7"},
     {:bumblebee, "~> 0.7"},
     {:tokenizers, "~> 0.5"},
     {:axon, "~> 0.8"},

livebooks/mnist_training.livemd

@@ -3,7 +3,7 @@
 ```elixir
 Mix.install(
   [
-    {:emily, "~> 0.6"},
+    {:emily, "~> 0.7"},
     {:axon, "~> 0.8"},
     {:scidata, "~> 0.1"},
     {:nx, "~> 0.12"},

livebooks/modernbert_classification.livemd

@@ -3,7 +3,7 @@
 ```elixir
 Mix.install(
   [
-    {:emily, "~> 0.6"},
+    {:emily, "~> 0.7"},
     {:bumblebee, "~> 0.7"},
     {:tokenizers, "~> 0.5"},
     {:nx, "~> 0.12"},

livebooks/nomic_embeddings.livemd

@@ -3,7 +3,7 @@
 ```elixir
 Mix.install(
   [
-    {:emily, "~> 0.6"},
+    {:emily, "~> 0.7"},
     {:bumblebee, "~> 0.7"},
     {:tokenizers, "~> 0.5"},
     {:nx, "~> 0.12"},

livebooks/qwen3_quantized.livemd

@@ -3,7 +3,7 @@
 ```elixir
 Mix.install(
   [
-    {:emily, "~> 0.6"},
+    {:emily, "~> 0.7"},
     {:bumblebee, "~> 0.7"},
     {:tokenizers, "~> 0.5"},
     {:nx, "~> 0.12"},

livebooks/smollm3_chat.livemd

@@ -3,7 +3,7 @@
 ```elixir
 Mix.install(
   [
-    {:emily, "~> 0.6"},
+    {:emily, "~> 0.7"},
     {:bumblebee, "~> 0.7"},
     {:tokenizers, "~> 0.5"},
     {:nx, "~> 0.12"},